Terms and Conditions

Version 1.0

Effective date: 4 August 2026

Last updated: 4 August 2026

These Terms are a legally binding agreement between you and PHASE 2 B.V. Please read them before you create an account. Section 19 contains a limitation of our liability. Section 22 tells you which law applies and where disputes are heard. If you are a consumer, Section 34 explains your right of withdrawal and how you can lose it.

PART 0 — PRELIMINARY

1. Scope of this document

1.1 What this document is. These Terms and Conditions (the "Terms") govern your access to and use of the Validating platform, in all of its parts: the marketing website at validating.studio, the web application at app.validating.studio, the Validating mobile applications for iOS and Android, the public application programming interface at api.validating.studio, and every other service, feature, content, tool, report, notification or communication that we make available in connection with them (together, the "Service").

1.2 Who this document binds. These Terms apply to every person who accesses or uses the Service, whether as a visitor, as a registered Ideator, as a registered Validator, as an administrator or member of a Workspace, as a developer using our API, or in any other capacity. Where a clause applies only to one category of user, it says so expressly.

1.3 How this document is structured. These Terms are organised in four parts. Part 0 and Part A apply to everyone. Part B applies additionally to Ideators. Part C applies additionally to Validators. Part D contains the annexes, which form an integral part of these Terms. If you use the Service in more than one capacity, more than one part applies to you.

1.4 Order of precedence. If there is a conflict between the parts of this document, the part that is specific to your role (Part B or Part C) prevails over the general part (Part A) in respect of the subject matter it governs. If there is a conflict between these Terms and a document incorporated by reference under Section 3, these Terms prevail, except where the incorporated document states expressly that it prevails or where the conflicting provision is required by law.

1.5 What this document is not. These Terms are not a privacy notice. Our processing of personal data is described in the Privacy Policy referred to in Section 3. These Terms are not investment, legal, tax, medical or business advice, and nothing produced by the Service constitutes such advice.

1.6 Acceptance. You accept these Terms by ticking the acceptance box presented at sign-up, by creating an account, or by otherwise accessing or using the Service. If you do not accept these Terms, you must not use the Service. If you accept these Terms on behalf of a company, organisation or other legal entity, you represent that you have authority to bind that entity, and "you" refers to that entity.

1.7 Record of the agreement. We store the version of these Terms you accepted, together with the date and time of acceptance and the identifier of your account. On request we will provide you with a copy on a durable medium. The Terms are concluded in English.

2. Who we are, and how to contact us

2.1 The provider of the Service.

• Legal name: PHASE 2 B.V.
• Legal form: Besloten vennootschap met beperkte aansprakelijkheid (Dutch private limited company)
• Registered office: De Boelelaan 1095a, Room 0A16, 1081 HV Amsterdam, the Netherlands
• Dutch Chamber of Commerce (KvK) number: 86210483
• VAT identification number: NL863896029B01
• General contact: hello@validating.studio

In these Terms, "Validating", "we", "us" and "our" mean the entity described above.

2.2 User contact point (Article 12 of Regulation (EU) 2022/2065). You may contact us about any matter relating to the Service, including any decision we have taken about your account or your content, at hello@validating.studio. We communicate in English and Italian. This contact point is not operated exclusively by automated means: you are entitled to reach a human being, and you may ask for a human at any point in the exchange.

2.3 Authorities contact point (Article 11 of Regulation (EU) 2022/2065). Member State authorities, the European Commission and the European Board for Digital Services may contact us at gabri@validating.studio. We accept communications in English and Italian.

2.4 Illegal content reports. Notices concerning allegedly illegal content are made through the mechanism described in Section 11 and at gabri@validating.studio.

2.5 Data protection contact. Requests concerning personal data are made to gabri@validating.studio. We have not appointed a Data Protection Officer, because our processing does not meet the conditions in Article 37 of Regulation (EU) 2016/679; we will appoint one and publish the contact details here if that changes.

2.6 Accessibility contact. Accessibility barriers may be reported to hello@validating.studio, as described in Section 15.

3. Documents incorporated by reference

3.1 Documents that form part of these Terms. The following documents are incorporated into and form an integral part of these Terms. By accepting these Terms you accept them too:

• Privacy Policy — What it covers: How we process personal data, on what legal basis, for how long, with whom we share it, and your rights · Where it lives: https://www.iubenda.com/privacy-policy/54962933
• Cookie Policy — What it covers: Cookies and similar technologies, and your choices · Where it lives: https://www.iubenda.com/privacy-policy/54962933/cookie-policy
• AI Transparency Notice — What it covers: Which parts of the Service use artificial intelligence, what those systems do, their known limitations, and how automated decisions are made and can be challenged · Where it lives: Part E of this document
• Acceptable Use Policy — What it covers: The detailed list of prohibited content and prohibited conduct (summarised in Section 8 and set out in full in Annex B) · Where it lives: Annex B to these Terms
• Instructions for Use of the AI systems — What it covers: The intended purpose, the conditions of use and the known limitations of the AI systems we make available · Where it lives: Part E of this document
• Fee Schedule — What it covers: The prices, fees, thresholds, minimums and limits in force from time to time · Where it lives: Displayed in the Service, in full, at the point of purchase
• Accessibility Statement — What it covers: The accessibility status of the Service and how to report barriers · Where it lives: Section 15 of this document

3.2 Documents that are not part of these Terms but govern part of your relationship. Some services are provided to you by third parties under their own terms, to which you agree separately and directly with them. We are not a party to those agreements. They include:

(a) the Stripe Connected Account Agreement and the other Stripe terms applicable to payments, payouts and identity verification (Sections 13 and 41);

(b) the Apple App Store Terms and Conditions and the Google Play Terms of Service, where you obtain or pay for anything through those stores (Section 16);

(c) the terms of any third-party integration you choose to connect, such as Slack or Notion (Section 32).

3.3 Changes to incorporated documents. We may amend the incorporated documents in accordance with Section 21. A material amendment is notified in the same way as a material amendment to these Terms.

3.4 Headings, examples and annexes. Headings are for convenience only and do not affect interpretation. Words introduced by "including", "such as" or "for example" are illustrative and not exhaustive. The annexes in Part D have the same force as the body of these Terms.

4. Definitions

In these Terms, the following capitalised words have the following meanings.

Account — the personal, non-transferable account you create in order to use the Service.

Agentic Panel — the evaluation of an Experiment by artificial intelligence agents rather than by human beings, as described in Section 9.4.

AI Output — any text, score, grade, ranking, verdict, summary, reasoning trace, recommendation, suggestion or other material generated wholly or predominantly by an artificial intelligence system operated by or for Validating, including Hypotheses, Reports, panel verdicts and quality scores.

Content Standards — the rules on permitted and prohibited content and conduct set out in Section 8 and Annex B.

Credits — a non-monetary balance that may be applied against the price of Experiments, as described in Section 29.

Experiment — a structured set of questions, tasks, materials and acceptance criteria created by an Ideator and submitted to a Panel in order to test a Hypothesis.

Human Panel — the evaluation of an Experiment by human Validators.

Hybrid Panel — the evaluation of the same Experiment by both a Human Panel and an Agentic Panel.

Hypothesis — the statement of the customer, problem and solution that an Experiment is designed to test.

Ideator — a user who creates, funds and runs Experiments. Part B applies to Ideators.

Offer — an invitation sent to a Validator to complete a specific Experiment for a stated amount, as described in Section 38.

Panel — a Human Panel, an Agentic Panel or a Hybrid Panel.

Report — the document generated at the end of an Experiment, containing the verdict, the underlying metrics and the AI-generated analysis.

Response — a Validator's completed set of answers to an Experiment, or the structured output of an artificial intelligence agent in an Agentic Panel.

Service — as defined in Section 1.1.

Terms — this document, including its annexes and the documents incorporated under Section 3.

User Content — any material you submit, upload, transmit, generate or store through the Service, including Hypotheses, Experiment questions and materials, mock-ups, prototypes, files, free-text answers, profile information, messages and reports of abuse. AI Output is not User Content, but it may incorporate or be derived from User Content.

Validator — a user who receives Offers and completes Experiments in exchange for payment. Part C applies to Validators.

Wallet — the record of amounts earned by a Validator and not yet paid out, as described in Section 40.

Workspace — a shared environment in which several users may access the same Experiments under a single subscription, as described in Section 33.

5. Who may use the Service

5.1 Intended audience. The Service is designed for two distinct audiences:

(a) Ideators — founders, product managers, consultants, researchers, students and other individuals or organisations who wish to test a business, product or service idea against structured feedback before building it. Ideators may act in a professional capacity or in a personal capacity; both are permitted, and the consumer protections in Section 34 and Section 22.5 apply to those who act outside their trade, business, craft or profession.

(b) Validators — individuals who wish to be paid for providing considered, honest feedback on ideas, on an occasional and entirely self-directed basis, from their own mobile device.

The Service is not designed for, and must not be used by, persons acting on behalf of an undisclosed third party, persons operating multiple identities, or automated agents purporting to be human beings.

5.2 Minimum age. You must be at least 18 years old to create an Account of any kind. We do not knowingly permit the Service to be used by persons below that age, and we do not knowingly collect personal data from them.

5.3 Higher age requirement for Validators. Because Validators receive payments, are subject to identity verification and assume their own tax obligations, a Validator must be at least 18 years old, without exception, and must satisfy any higher minimum age applicable in their country of residence.

5.4 Age verification and consequences. We may verify your age at any time, including through the identity verification described in Section 41. If we establish or reasonably believe that you are below the applicable minimum age we will close your Account. Where the Account is a Validator Account, amounts genuinely earned before closure will be paid out to the extent this is lawful and technically possible, and the remaining personal data will be deleted or anonymised.

5.5 Capacity. You must have the legal capacity to enter into a binding contract. If you are subject to a measure that restricts your legal capacity, you may use the Service only to the extent permitted by that measure.

5.6 Eligible countries. The Service is offered only in the countries shown as available during sign-up, and the current list may be obtained at any time from hello@validating.studio. Availability may differ between the Ideator side and the Validator side of the Service, because the ability to receive payouts depends on the coverage of our payment provider and on local law. We may add or remove countries; if we remove the country in which you are established or resident, Section 14.6 applies.

5.7 Sanctions and restricted persons. You may not use the Service if you are located in, ordinarily resident in, or organised under the laws of a country or territory subject to comprehensive economic sanctions of the European Union, the Netherlands, the United Kingdom or the United States, or if you appear on any applicable list of restricted or designated persons. You represent that neither you nor, where you act for an entity, that entity or any person controlling it, is such a person.

5.8 One person, one Account. You may hold one Ideator Account and one Validator Account. You may not hold more than one Account of the same type, and you may not create an Account on behalf of another person, use another person's identity or credentials, or allow another person to use yours. Where an organisation needs several people to work on the same Experiments, the Workspace feature described in Section 33 is the permitted mechanism.

5.9 Ideator and Validator at the same time. You may hold both an Ideator Account and a Validator Account, provided that you never act as a Validator in relation to an Experiment created by you, by a person connected to you, or by an organisation in which you have an interest. Doing so is a material breach of these Terms.

6. Registration, authentication and account security

6.1 Creating an Account. You create an Account by signing in with a supported identity provider (currently Google, Apple and LinkedIn) or by requesting a single-use sign-in link sent to your email address. We do not use passwords. At sign-up you choose whether you are registering as an Ideator or as a Validator.

6.2 Choice of role. The role you choose determines the entire experience that follows, including which application you use and which part of these Terms applies to you. The role of an existing Account cannot currently be changed by you. If you have chosen the wrong role, contact us at hello@validating.studio and we will assist you, which may involve closing the Account and opening a new one.

6.3 Accuracy of the information you give us. The information you provide during registration and profile completion must be true, accurate, current and complete, and you must keep it up to date. This obligation is not merely formal: the Service matches Experiments to Validators on the basis of the professional and demographic information they declare, and the value of the Service to Ideators depends on that information being truthful. Deliberately false profile information is a material breach of these Terms.

6.4 Sign-in links and account security. Sign-in links are personal, single-use and expire. You must not share them, forward them or publish them. You are responsible for keeping your email account, your identity provider account and your device secure, and for all activity that takes place through your Account. You must notify us without undue delay at hello@validating.studio if you know or suspect that your Account has been accessed by anyone else.

6.5 Sessions and devices. You may be signed in on the web application and the mobile application at the same time. Sessions expire and must be renewed. We may terminate sessions for security reasons.

6.6 Validator onboarding. Before a Validator Account can receive Offers, you must complete the onboarding described in Section 37, and, at the point described in Section 41, identity verification. We may accept, refuse or defer any application to become a Validator, as described in Section 37.4.

6.7 Administrative access. Members of our support and trust-and-safety staff may, where necessary to investigate a report, resolve a support request or comply with a legal obligation, view your Account in a read-only mode that reproduces what you see. Such access is limited in time, is logged with the identity of the person accessing, the time and the reason, and does not permit any action to be taken in your name. Where such access has taken place other than at your own request, we will inform you on request.

PART A — PROVISIONS COMMON TO ALL USERS

7. What the Service does, and what it does not do

7.1 A two-sided marketplace. Validating operates a marketplace with two sides. On one side, Ideators describe an idea, define a Hypothesis, set a budget and launch an Experiment. On the other side, Validators receive Offers to complete Experiments matched to their declared profile and are paid for the Responses they submit. Validating provides the platform, the matching, the payment infrastructure, the artificial intelligence layer and the reporting.

7.2 Three types of Panel. An Ideator chooses, when creating an Experiment, whether it is submitted to a Human Panel, to an Agentic Panel, or to both (a Hybrid Panel). The characteristics, price, speed and limitations of each are displayed at the point of choice and are described in Section 9.

7.3 Our role in relation to Responses. We provide the infrastructure that collects, structures, aggregates and analyses Responses. We do not warrant the truthfulness, accuracy, completeness, representativeness or commercial value of any Response, of any Report, or of any verdict. Validators are independent individuals who express their own opinions. Artificial intelligence agents produce outputs that reflect the model's training and prompting, not the behaviour of any real person. Section 18 sets out the disclaimers in full.

7.4 We do not act as an agent. We are not the agent, representative, employer, partner or joint venturer of any Ideator or any Validator. We contract with each of them in our own name.

7.5 What is not part of the Service. Unless we state otherwise in writing, the Service does not include: market research services regulated as such under any professional standard; recruitment, employment or staffing services; clinical, medical or psychological research; investment research or financial promotion; legal, accounting or tax advice; or any service requiring a licence we do not hold. You must not use the Service to obtain any of them.

7.6 Beta and preview features. We may make features available on a beta, preview, early-access or experimental basis. Such features are identified as such, are provided "as is", may be changed or withdrawn at any time without notice, may not be covered by our support commitments, and are excluded from any service level described in these Terms.

7.7 The Service evolves. The Service is under active development. Features described in our documentation as planned, forthcoming or applicable to a future version do not form part of the Service until we make them available. Nothing in our marketing communications, roadmaps or public statements creates a contractual obligation to deliver a feature.

7.8 Advertising. The Service currently displays no third-party advertising, and Experiments are not sold as advertising space. If we introduce advertising or sponsored placements in the future:

(a) every advertisement will be identifiable as such, in a clear, concise and unambiguous way, together with the identity of the person on whose behalf it is presented;

(b) we will not present a sponsored Experiment, a sponsored result, or a promoted position as an organic one;

(c) we will not target advertising on the basis of special categories of personal data, and we will not target advertising to a person we know or reasonably believe to be a minor;

(d) we will disclose the main parameters used to determine to whom an advertisement is shown, and the means of changing them.

7.9 Ranking and prioritisation. Where the order or visibility of Offers, results or listings is determined by us, the main parameters that determine it, and their relative importance, are described in Section 42.1 and in Part E. Where a payment influences prioritisation — for example where an Ideator purchases accelerated distribution — the fact and the effect of that payment are disclosed at the point of use.

7.10 Commercial communications. We send you transactional messages that are necessary to operate the Service, such as sign-in links, receipts, Offers, results and legal notices; these are not marketing and you cannot opt out of them while you hold an Account. Marketing messages are sent only where you have consented or where the law otherwise permits, and every one of them contains a one-click means of unsubscribing that we honour without delay.

8. Acceptable use

8.1 The general rule. You may use the Service only for lawful purposes, in accordance with these Terms, and in a manner that respects the rights, safety and dignity of others. Annex B sets out the Content Standards in full and forms part of these Terms.

8.2 Prohibited content — summary. You must not submit, upload, transmit, generate, request or distribute through the Service any content that:

(a) is unlawful under Union law or under the law of any Member State or other country where it is accessible, or that promotes, facilitates or instructs on unlawful activity;

(b) constitutes child sexual abuse material, sexualises a minor in any way, or is a non-consensual intimate image, including any synthetic or manipulated image of that nature;

(c) is defamatory, harassing, threatening, abusive, or that incites violence, hatred or discrimination against a person or group on the basis of a protected characteristic;

(d) infringes the intellectual property rights, trade secrets, privacy rights, image rights or confidentiality rights of any person;

(e) contains the personal data of a third party without a lawful basis, or contains special categories of personal data of any person outside the framework described in Section 10.6;

(f) contains malicious code, or a link to it, or is designed to interfere with the operation of the Service or of any device;

(g) is deceptive as to its origin, purpose or authorship, including content presented as human-authored that was generated by an artificial intelligence system in a context where that would mislead;

(h) is designed to test, promote or develop a product or activity that is itself illegal, or the marketing of which is prohibited, in the market to which the Experiment relates;

(i) is gratuitously shocking, obscene or distressing, having regard to the fact that Validators do not choose the content of the Experiments they are offered beyond the broad categories in their profile.

8.3 Prohibited conduct — summary. You must not:

(a) use the Service to build, train, fine-tune, evaluate or benchmark a machine-learning model, or to create a dataset for any of those purposes, whether by scraping, by automated collection, or by systematically extracting Responses, Reports or reasoning traces;

(b) reverse engineer, decompile, disassemble or attempt to derive the source code, prompts, model weights, matching logic, scoring logic or algorithms of the Service, except to the strict extent that such an act cannot lawfully be prohibited;

(c) access the Service by automated means other than the API described in Section 31, or circumvent any rate limit, access control, authentication mechanism or technical protection measure;

(d) hold more than one Account of the same type, impersonate any person, or misrepresent your identity, profession, experience, location or affiliation;

(e) act as a Validator in relation to your own Experiment, or coordinate with others to influence the outcome of an Experiment;

(f) submit Responses generated by an automated tool, submit Responses without genuinely engaging with the material, or otherwise game any quality, trust or matching mechanism;

(g) resell, sublicense, rent, lease or otherwise make the Service available to any third party, except within a Workspace as permitted by Section 33;

(h) remove, alter, obscure or degrade any provenance marking, watermark, metadata, label or notice that identifies content as artificially generated, or as originating from the Service;

(i) use the Service in a manner that imposes an unreasonable or disproportionate load on our infrastructure, or that interferes with the use of the Service by others;

(j) use the Service for any of the practices prohibited by Article 5 of Regulation (EU) 2024/1689.

8.4 Confidentiality of what you see. Experiments frequently contain unpublished ideas, unreleased product concepts, mock-ups and prototypes. If you see such material through the Service — in particular as a Validator, or as an administrator of a Workspace — you must treat it as confidential. You must not copy it, record it, screenshot it, republish it, disclose it to any third party, or use it for any purpose other than completing the Experiment or performing your role. This obligation survives the closure of your Account and continues for so long as the material is not public through no act of yours. Section 45 develops this obligation for Validators.

8.5 Consequences. Breach of this Section, of Annex B or of the Content Standards may result in any of the measures described in Section 11.4, up to and including the immediate termination of your Account and, where applicable, the withholding of amounts as permitted by Section 40.10. We report to the competent authorities any content or conduct that gives rise to a suspicion of a criminal offence involving a threat to the life or safety of a person, as required by Article 18 of Regulation (EU) 2022/2065.

9. Artificial intelligence

9.1 Transparency — you are dealing with AI. Substantial parts of the Service are operated by artificial intelligence systems and not by human beings. The AI Transparency Notice referred to in Section 3 describes each of them, and each is labelled in the interface at the point of use. In summary, artificial intelligence is used to:

(a) generate follow-up questions and draft your Hypothesis from what you describe;

(b) score and comment on your Hypothesis, identify risks in it, and suggest improvements;

(c) draft the initial structure of an Experiment and its acceptance criteria;

(d) analyse free-text answers, including sentiment analysis, and compute the metrics on which a verdict depends;

(e) operate the Agentic Panel, in which artificial intelligence agents evaluate a Hypothesis and a solution and produce a grade and a reasoning trace;

(f) generate the Report, the insights and the recommendations;

(g) assess the quality of a Validator's submitted profile materials and of their Responses, and contribute to the matching of Experiments to Validators.

9.2 Artificially generated content is marked. AI Output is identified as artificially generated, both visibly in the interface and, where technically feasible, in a machine-readable form embedded in or accompanying the file. You must not remove, alter, obscure or degrade those markings. Doing so is a breach of Section 8.3(h) and may also breach Article 50 of Regulation (EU) 2024/1689.

9.3 The limits of AI Output — read this carefully. AI Output is generated by statistical models. It can be wrong. It can be confidently wrong. Specifically, and without limiting Section 18:

(a) AI Output may contain factual errors, invented references, misattributed statements and internally inconsistent reasoning;

(b) a verdict, grade or score is the mechanical result of applying the acceptance criteria that were configured for that Experiment to the Responses actually collected. It is not a prediction of commercial success, of market demand, or of the behaviour of any real population;

(c) the number and composition of the Panel are limited. Results are not statistically representative of any market, population or segment unless the Experiment was specifically designed and sized to be so, which the standard Service does not do;

(d) reasoning traces show what a model produced, not why it "really" decided anything. They are a description generated by the model, not an audited explanation of its internal state;

(e) AI Output does not constitute professional, legal, financial, medical, investment or regulatory advice, and must not be relied on as such.

9.4 What the Agentic Panel is, and what it is not. The Agentic Panel observes how large language models and agent systems reason about, include, compare, recommend or reject a product for a given hypothesis. It is not a synthetic audience. It does not simulate human beings. It does not claim to predict human behaviour. An Agentic Panel result tells you how a machine evaluated your idea. It tells you nothing, of itself, about how a person would.

9.5 You remain responsible for your decisions. You are solely responsible for any decision you take on the basis of AI Output, including any decision to build, not to build, to invest, not to invest, to pivot, to price, to launch or to abandon. You must exercise your own judgement and, where the decision matters, obtain independent professional advice. Human oversight of the output is your responsibility as the person who uses it.

9.6 Roles under Regulation (EU) 2024/1689. Validating acts as a provider of the artificial intelligence systems that it makes available under its own name through the Service. Where you use the Service, you act as a deployer of those systems in respect of your own use of the output. Accordingly:

(a) you must use the systems in accordance with the Instructions for Use referred to in Section 3, and not for any purpose they do not cover;

(b) you must not substantially modify a system, nor put it into service under your own name or trade mark, nor change its intended purpose; if you do, you may become a provider in your own right under Article 25 of that Regulation, with all the obligations that entails, and Section 20 applies;

(c) if you make AI Output public, you are responsible for any transparency obligation that attaches to that publication, including Article 50(4) of that Regulation, unless you have subjected the output to human review and you assume editorial responsibility for it;

(d) you must preserve the markings referred to in Section 9.2.

9.7 Automated decisions about you. Some decisions taken by the Service are taken by automated means, without human intervention at the moment they are taken. They are:

• Matching — What it does: Determines which Validators receive an Offer for a given Experiment, and in what order · Who it affects: Validators
• Quality assessment — What it does: Scores the quality and internal consistency of the materials submitted during Validator onboarding, and of submitted Responses · Who it affects: Validators
• Trust score — What it does: Aggregates a Validator's history into a reliability figure that governs access to higher-paying Experiments · Who it affects: Validators
• Matching restriction — What it does: Reduces or suspends the Offers a Validator receives · Who it affects: Validators
• Fraud and abuse detection — What it does: Flags Accounts, Experiments or Responses for review · Who it affects: All users
• Verdict computation — What it does: Computes the grade and verdict of an Experiment from the configured acceptance criteria · Who it affects: Ideators

9.8 Your rights in respect of automated decisions. Where a decision listed in Section 9.7 produces legal effects concerning you or similarly significantly affects you, you have the right to:

(a) obtain human intervention — a qualified person at Validating reviews the decision;

(b) express your point of view — you may submit any explanation or evidence you consider relevant;

(c) obtain an explanation — we will describe, in plain language, the procedure and the principles actually applied, the main categories of data used and their relative importance, and the main reasons for the outcome in your case. A claim of trade secrecy does not entitle us to refuse an explanation altogether; where genuine confidentiality prevents full disclosure to you, we will provide the withheld information to the competent supervisory authority or court instead;

(d) contest the decision and obtain its rectification if it was wrong.

9.9 How to exercise those rights, and our deadline. Write to gabri@validating.studio stating the decision you are contesting. We will complete the human review and reply, with individual reasons, within 14 days of your request. No permanent suspension of an Account, and no forfeiture of compensation already earned, is imposed by automated means alone: a human being reviews it before it becomes final, and in any event on your request.

9.10 What we do not do with AI. We do not use artificial intelligence to infer your emotional or psychological state; to analyse your private communications; to predict whether you will exercise collective rights or join a trade union; or to infer racial or ethnic origin, migration status, political opinions, religious or philosophical beliefs, disability, health, or sex life. We do not use biometric data for any purpose other than the one-to-one identity verification described in Section 41, which is carried out by our payment provider.

9.11 Use of your material to improve models. We do not provide User Content to any third-party model provider for the purpose of training that provider's models, and we have contracted with our artificial intelligence suppliers on the basis that content submitted through their interfaces is not used to train their models. Where we use material generated through the Service to improve our own prompts, heuristics, evaluation criteria and quality controls, we do so on the basis set out in the Privacy Policy, in aggregated or de-identified form wherever that is sufficient for the purpose, and you may object as described there.

9.12 Aggregated insight and benchmarks. We create aggregated, de-identified statistics and benchmarks from Experiments and Responses — for example, the proportion of Experiments in a category that reach a given verdict, or recurring patterns in how machine agents evaluate a category. We may use those aggregates to operate, improve and market the Service, and to make comparative insight available to other users. Aggregates never identify you, your Experiment, your Hypothesis, your solution or any individual Response, and are constructed so that they cannot reasonably be re-identified. Section 10.2 grants us the licence needed to do this.

9.13 AI literacy. We take measures to support an adequate level of artificial intelligence literacy among our staff and among the users of the Service, including the AI Transparency Notice, the in-product explanations of what each system does and does not do, and the Instructions for Use.

10. Content, intellectual property and licences

Your content

10.1 You keep your rights. As between you and us, you retain all rights in your User Content. Nothing in these Terms transfers ownership of your ideas, your business concepts, your designs, your mock-ups, your prototypes or your text to us.

10.2 The licence you grant us. You grant us a worldwide, non-exclusive, royalty-free, transferable and sublicensable (to our hosting, artificial intelligence, payment, analytics and support providers, solely for the purposes below) licence to host, store, reproduce, transmit, display, adapt, reformat, translate, excerpt, analyse and create derived material from your User Content, for the following purposes only:

(a) operating, delivering and supporting the Service, including presenting your Experiment to Validators and to artificial intelligence agents;

(b) generating AI Output, including Hypotheses, analyses, verdicts and Reports;

(c) producing the aggregated, de-identified insight described in Section 9.12;

(d) enforcing these Terms, investigating reports and preventing fraud and abuse;

(e) complying with a legal obligation or responding to a lawful request from an authority.

This licence lasts as long as we host the content, and survives thereafter only to the extent necessary for purposes (c), (d) and (e), and for backups, until deletion in the ordinary course. It does not permit us to publish your Hypothesis, your solution or your uploaded materials to the public.

10.3 Your warranties about your content. You represent and warrant that you own or have all necessary rights, licences, consents and permissions in respect of your User Content; that its use as contemplated by these Terms does not infringe the rights of any third party or breach any obligation of confidence; that it complies with the Content Standards; and that any personal data of a third party it contains is included on a lawful basis and with the information required by data protection law having been given to that person.

10.4 Content you should not upload. Do not upload material that you cannot afford to have seen. Do not upload trade secrets belonging to a third party, material subject to an obligation of confidence you owe to someone else, unredacted personal data, credentials, financial account details, health information, or anything whose disclosure would cause serious harm. Experiments are shown to Validators, who are independent individuals bound by Section 8.4 and Section 30 but over whom we exercise no physical control.

10.5 Free-text answers. If you are a Validator, do not include in a free-text answer any information that identifies you or any third party. Ideators receive free-text answers under a pseudonymous label, but we cannot remove identifying information that you yourself choose to write into the body of an answer.

10.6 Special categories of data. An Experiment may not be designed to elicit special categories of personal data within the meaning of Article 9 of Regulation (EU) 2016/679 — data revealing racial or ethnic origin, political opinions, religious or philosophical beliefs, trade union membership, genetic or biometric data, data concerning health, sex life or sexual orientation — unless we have expressly approved it in advance in writing, the purpose is legitimate and clearly explained, participation is optional, and the Validator's explicit, specific and revocable consent is collected for that Experiment. Any Experiment that seeks such data without approval will be rejected, and the Ideator refunded, under Section 11.

Our content

10.7 Our intellectual property. The Service, and everything in it other than User Content — including the software, the interfaces, the design system, the block library, the matching logic, the prompt architecture, the evaluation criteria catalogue, the databases, the documentation, the name "Validating", our logo and our other marks — is owned by us or our licensors and is protected by intellectual property law. Nothing in these Terms transfers any of it to you.

10.8 Your licence to use the Service. Subject to your compliance with these Terms and to payment of any applicable fees, we grant you a limited, personal, non-exclusive, non-transferable, non-sublicensable and revocable right to access and use the Service for its intended purpose, for the duration of your Account. Every right not expressly granted is reserved.

10.9 Mobile application licence. The Validating mobile applications are licensed, not sold. We grant you a limited, personal, non-exclusive, non-transferable, revocable licence to install and run one copy of the application on a device you own or control, and to use it for its intended purpose, subject to the terms of the store from which you obtained it. You may not copy, modify, translate, merge, adapt, decompile, disassemble, reverse engineer or create derivative works of the application, nor circumvent any technical protection measure, except to the strict extent that such acts cannot lawfully be prohibited. Store-specific terms are set out in Section 16.

10.10 Open source. The Service incorporates third-party open-source components, each licensed under its own terms. Where an open-source licence conflicts with this Section, that licence prevails in respect of the component it covers. A list is available on request.

Rights in the output

10.11 Rights in the Report. To the extent that any intellectual property right subsists in a Report generated for your Experiment, and to the extent that such right is capable of assignment, we assign it to you on payment in full for the Experiment. To the extent that it is not capable of assignment, or does not subsist, we grant you a worldwide, perpetual, irrevocable, transferable and sublicensable licence to use, reproduce, adapt and communicate that Report for any purpose.

10.12 A warning about the protectability of AI Output. Under Union law, copyright protection requires a human creative contribution. A Report or other AI Output that is generated predominantly by an automated system may not attract copyright protection at all, in your hands or in ours. We make no representation and give no warranty as to whether any AI Output is protectable, in any jurisdiction. If protection matters to you, take advice before you rely on it.

10.13 Outputs are not unique to you. Artificial intelligence systems generate similar outputs from similar inputs. Another user may receive substantively similar AI Output. Section 10.11 grants you rights in your Report; it does not grant you exclusivity over ideas, formulations or analyses that others may independently receive.

10.14 Validation certificates. Where the Service allows you to generate a public certificate recording the outcome of an Experiment, you may publish and share it. You are responsible for the accuracy of any claim you make on the basis of it, and for ensuring that your use of it is not misleading under consumer or advertising law. A certificate records that an Experiment reached a given verdict on a given date with a given number of participants. It is not an endorsement, a certification, an accreditation, or a statement by us about the merits of your product. We may revoke a certificate that relates to an Experiment obtained in breach of these Terms.

10.15 Feedback. If you send us suggestions, ideas or feedback about the Service, you grant us a perpetual, irrevocable, worldwide, royalty-free right to use them without restriction and without any obligation to you. Do not send us anything you wish to keep or to be paid for.

11. Reports of illegal content, moderation, and complaints

11.1 Prior review of Experiments. Every Experiment submitted by an Ideator is reviewed before it is distributed to a Human Panel. The review checks that the Experiment complies with the Content Standards and is intelligible and functional. The outcome is either approval and launch, or rejection with a refund under Section 28.4. Our target is to complete this review within four hours, which is an operational objective and not a contractual commitment.

11.2 Notice and action. Anyone — whether or not a user — may notify us of content on the Service that they consider illegal, by writing to gabri@validating.studio. To enable us to act, a notice should contain: a sufficiently substantiated explanation of why the content is illegal; the exact electronic location of the content, such as the URL or the Experiment identifier; the name and email address of the person submitting the notice, except where the notice concerns an offence referred to in Articles 3 to 7 of Directive 2011/93/EU; and a statement that the person submitting the notice believes in good faith that the information is accurate and complete.

11.3 What we do with a notice. We confirm receipt without undue delay. We process notices in a timely, diligent, non-arbitrary and objective manner. Where a notice allows us to establish the illegality of the content without a detailed legal examination, we act expeditiously. We notify the person who submitted the notice of our decision and of the redress available against it. Where we use automated means to process or decide on a notice, we say so.

11.4 Measures we may take. Where content or conduct breaches the Content Standards or the law, we may, proportionately to the seriousness of the breach and having regard to the fundamental rights of everyone concerned:

(a) remove the content, disable access to it, or restrict its visibility;

(b) refuse to launch an Experiment, or terminate a running Experiment;

(c) reduce or suspend the Offers a Validator receives;

(d) suspend or terminate the provision of the Service to you, in whole or in part;

(e) suspend or terminate your Account;

(f) suspend a payment, in accordance with Section 40.10;

(g) demote or delist content, refuse to generate a certificate, or revoke one.

11.5 Statement of reasons. Whenever we take a measure listed in Section 11.4 against you, we will give you a clear and specific statement of reasons, which will set out: the measure taken, its territorial and temporal scope; the facts and circumstances relied on; whether automated means were used to detect or to decide, and if so which; the contractual ground or the legal ground relied on, with an explanation of why the content or conduct is incompatible with it; and the redress available to you, including internal complaint, out-of-court dispute settlement where applicable, and judicial redress. We will provide the statement of reasons at the latest when the measure takes effect, unless the measure is required by an order of an authority which asks us not to disclose it. Where we are required to do so, we submit the statement of reasons to the Commission's transparency database.

11.6 Internal complaint. You may complain against any measure by writing to gabri@validating.studio within six months of being notified of it. Complaints are handled by qualified staff, are not decided solely by automated means, and are answered with reasons. We aim to answer within 14 days and will not exceed 30 days. If we uphold your complaint we reverse the measure without undue delay and restore what can be restored.

11.7 Misuse. We may suspend, after having issued a prior warning, the provision of the Service to users who frequently submit manifestly illegal content, and the processing of notices and complaints from persons or entities that frequently submit manifestly unfounded ones. Whether conduct is frequent and manifestly unfounded is assessed case by case, in a diligent, objective and proportionate manner, having regard to the number and the ratio of such submissions and to their gravity.

11.8 Out-of-court dispute settlement and judicial redress. Nothing in this Section deprives you of the right to bring proceedings before a court, or, where you are a consumer, of your right to use the out-of-court dispute settlement mechanisms referred to in Section 22.6.

11.9 Orders from authorities. Where we receive an order from a competent judicial or administrative authority to act against specific illegal content, or to provide specific information, we comply and inform the authority of the effect given to the order. Where the order concerns you and the law permits, we inform you of the order and of the effect given to it, at the latest when the effect is given or at the time specified by the authority.

12. Personal data

12.1 The Privacy Policy governs. How we collect, use, share, transfer and retain personal data is described in the Privacy Policy referred to in Section 3, which forms part of these Terms. This Section states only the points that bear directly on the contract.

12.2 Our role. In respect of the personal data of Ideators and Validators — account data, profile data, professional and demographic attributes, payment and payout data, Responses, usage data — Validating acts as controller. We determine the purposes and means of that processing, and we are accountable for it.

12.3 Personal data inside your content. Where an Ideator includes personal data of a third party in the material uploaded to an Experiment, the Ideator is responsible for having a lawful basis for doing so and for having given that person the information required by law. We process such data on the Ideator's behalf, as processor, solely to deliver the Service. Where processing of that kind takes place, our standard data processing terms apply and are incorporated by reference; a copy is available on request from gabri@validating.studio. If no such terms have been agreed with you, the Ideator must not include third-party personal data in Experiment material.

12.4 What Ideators never see. Ideators do not receive the identity of any Validator. Responses are presented under a pseudonymous label. Ideators do not receive the raw values of sensitive profile attributes, including income, exact age, gender and healthcare qualifications; they receive only anonymised match criteria. Inferred behavioural scores and trust scores are internal and are never disclosed to Ideators or to any third party.

12.5 Sub-processors. We use third-party providers to host, process, analyse, secure, communicate and pay. They include our database and storage provider, our hosting provider, our artificial intelligence provider, our payment and identity provider, our transactional email provider, our push notification provider, our analytics provider and our error-monitoring provider. The current list, with each provider's location and role, is maintained in the Privacy Policy. Where a provider is outside the European Economic Area, transfers are made under an adequacy decision or under appropriate safeguards, with a transfer impact assessment on file.

12.6 Data location and retention. Our primary database and storage are hosted within the European Union. Raw Responses are deleted two years after the completion of the Experiment to which they relate. Aggregated and anonymised data is retained indefinitely, in a form that does not permit re-identification. Other retention periods are stated in the Privacy Policy.

12.7 Your rights. You may exercise your rights of access, rectification, erasure, restriction, portability and objection, and your right to withdraw a consent, by writing to gabri@validating.studio or through the in-product controls. We respond within the time limits set by law. Deleting a Validator profile anonymises the personal data while preserving aggregated statistics that no longer identify you. Deleting your Account does not, by itself, cancel a running Experiment, a subscription, or an obligation to pay; and it does not forfeit a Wallet balance you have already earned — Section 40.11 explains what happens to it.

12.8 Impact assessment. We maintain a data protection impact assessment covering the automated processing described in Section 9.7 and keep it under review.

12.9 Security. We apply technical and organisational measures appropriate to the risk, including encryption in transit and at rest, row-level access control on every database table, no storage of passwords, restricted and logged administrative access, regular backups with tested restoration, and independent security testing before production release. No system is perfectly secure. You must also play your part, in accordance with Section 6.4.

13. Payments infrastructure, and the role of our payment provider

13.1 We do not handle your card. All card payments, all payouts and all identity verification are processed by Stripe and its affiliates. We do not receive, store or process full payment card numbers, and we are not within the scope of the card industry security standards for that reason.

13.2 Your agreement with Stripe. By making a payment or by receiving a payout through the Service you also enter into an agreement directly with Stripe, including the Stripe Connected Account Agreement where you receive payouts. That agreement is between you and Stripe. You are responsible for complying with it, and Stripe may impose its own verification, information and eligibility requirements. If Stripe refuses, restricts or terminates your relationship with it, we may be unable to provide the payment or payout parts of the Service to you, and Section 17 applies.

13.3 Payment method. You must use a payment method you are authorised to use. You authorise us and Stripe to charge that method for the amounts described in Part B, including recurring subscription charges, and to place authorisations on it in anticipation of a charge.

13.4 Chargebacks and disputes. If you dispute a charge with your card issuer, we may suspend the affected Experiment, the resulting Report and your Account while the dispute is resolved, and we may recover from you the amount of the disputed charge together with any fee imposed on us in connection with it, where the dispute is resolved in our favour or where the charge was legitimate. Contact us at hello@validating.studio before raising a chargeback; we would rather resolve it directly.

13.5 Currency. Prices, budgets, fees, Wallet balances and payouts are denominated in euro (EUR) unless the interface states otherwise for your country. Where your bank or card issuer converts a currency, its exchange rate and any conversion fee are a matter between you and it. Any conversion fee applied by our payment provider on a payout is disclosed in the Fee Schedule.

13.6 Set-off. We may set off any amount you owe us under these Terms against any amount we owe you, including a Wallet balance, to the extent permitted by applicable law and, where you are a consumer, only in respect of an undisputed or judicially determined claim.

14. Availability, maintenance and changes to the Service

14.1 We aim high but we do not promise perfection. We aim to make the Service available at all times, and we publish internal availability objectives. Those objectives are engineering targets, not contractual service levels, and no service credit, penalty or remedy attaches to a failure to meet them, save for your rights as a consumer under Section 22.5 and Section 18.6.

14.2 Planned maintenance. We carry out planned maintenance in a recurring window, currently Tuesdays between 02:00 and 04:00 Central European Time, announced at least 48 hours in advance where the maintenance is expected to interrupt the Service.

14.3 Unplanned interruption. The Service may be unavailable, degraded or interrupted because of a fault, a security incident, a failure of a third-party provider, an act of a third party, or a circumstance beyond our control. We will restore the Service as promptly as we reasonably can.

14.4 Emergency measures. We may suspend the Service or any part of it, without notice, where necessary to address a security threat, to prevent harm to users, to comply with a legal obligation, or to preserve the integrity of the Service. We will give notice as soon as we reasonably can.

14.5 Changes to the Service. We may add, change, improve, restrict or remove features. Where a change materially reduces the functionality of a paid part of the Service that you are currently paying for, we will notify you at least 30 days in advance, and you may terminate the affected paid part with effect from the date the change takes effect and receive a refund of the unused portion of any amount paid in advance for it. Changes that are necessary for security, legality or the correction of a defect may take effect immediately.

14.6 Withdrawal of the Service in a country. If we cease to offer the Service in a country, we will give affected users at least 30 days' notice where circumstances permit, allow them to export their data under Section 35, complete or refund Experiments in progress, and pay out Wallet balances.

14.7 Discontinuation of the Service. If we decide to discontinue the Service altogether, we will give at least 90 days' notice, allow export of data throughout that period, refund the unused portion of any prepaid amount, complete or refund Experiments in progress, and pay out Wallet balances in accordance with Section 40.

14.8 Third-party dependencies. Parts of the Service depend on third parties, including artificial intelligence model providers, payment providers, identity providers, notification providers and hosting providers. If such a provider changes, restricts or discontinues its service, we may have to change ours, and Section 14.5 applies.

15. Accessibility

15.1 Our commitment. We design the Service to be usable by as many people as possible. Our target is conformity with EN 301 549 and, in practice, with the Web Content Accessibility Guidelines at level AA.

15.2 Statement. PHASE 2 B.V. is at present a microenterprise within the meaning of Article 4(5) of Directive (EU) 2019/882 — fewer than ten persons employed and an annual turnover and balance sheet total not exceeding two million euro — and as a service provider is therefore exempt from the accessibility requirements of that Directive. We do not treat the exemption as a reason to build an inaccessible product: we design to the standard stated in Section 15.1, and we will publish a full Accessibility Statement, and comply in full, from the moment we cease to qualify for the exemption.

15.3 Accessible AI information. Regardless of any exemption, the information we give you about the use of artificial intelligence under Section 9 is provided in an accessible format.

15.4 Reporting a barrier. If any part of the Service is inaccessible to you, write to hello@validating.studio. Tell us what you were trying to do, what stopped you and what assistive technology you use. We aim to respond within 10 working days and to offer an alternative means of completing the task in the meantime.

16. Mobile applications and app stores

16.1 Where the applications come from. The Validating mobile applications are distributed through the Apple App Store and the Google Play Store. Your acquisition and use of an application is also subject to the terms of the store you obtained it from.

16.2 Device requirements. The applications require a supported operating system version and a minimum device specification, published in the store listing. We may cease to support older versions; where we do, we will give notice in the application.

16.3 Updates. We may issue updates, which may be installed automatically depending on your device settings. We may require you to install an update in order to continue using the application, in particular where the update addresses a security issue or a legal requirement.

16.4 Notifications. The Validator application relies on push notifications to deliver Offers. If you disable them, you will receive fewer Offers or none, and we are not responsible for Offers you do not see. Notification permission is requested in accordance with platform rules and may be withdrawn at any time in your device settings.

16.5 Offline use and local storage. The application may store data on your device, including for offline completion of an Experiment. Data stored locally is synchronised when connectivity returns. If you delete the application before synchronisation, unsynchronised work may be lost.

16.6 Apple-specific terms. Where you obtained the application from the Apple App Store, the following applies and prevails over any conflicting provision of these Terms in respect of that application: these Terms are between you and us only, and not with Apple; we, and not Apple, are solely responsible for the application and its content; Apple has no obligation to furnish maintenance or support; in the event of a failure of the application to conform to any applicable warranty, you may notify Apple, which will refund the purchase price, and Apple has no other warranty obligation; we, and not Apple, are responsible for addressing any claim by you or a third party relating to the application, including product liability, failure to conform to legal requirements, and consumer protection claims; we, and not Apple, are responsible for the investigation, defence, settlement and discharge of any third-party claim that the application infringes intellectual property rights; you represent that you are not located in a country subject to a United States Government embargo or designated as a terrorist-supporting country, and that you are not on any United States Government list of prohibited or restricted parties; and Apple and its subsidiaries are third-party beneficiaries of these Terms in respect of the application and may enforce them against you.

16.7 Google-specific terms. Where you obtained the application from the Google Play Store, your use is subject to the Google Play Terms of Service, and Google is not a party to these Terms and has no responsibility for the application.

16.8 Purchases made through a store. Where a subscription is sold through an app store, the store is the merchant of record. Billing, renewal, cancellation and refunds for that purchase are governed by the store's rules, and you must exercise cancellation and refund rights through the store. We cannot refund a store purchase directly.

17. Suspension, termination and the consequences

17.1 You may leave at any time. You may close your Account at any time by writing to hello@validating.studio or, where available, through the in-product control. Closure takes effect once any Experiment in progress has been completed or cancelled and any amount due has been settled.

17.2 When we may suspend or terminate. We may suspend or terminate your Account, or any part of the Service provided to you, where:

(a) you are in material breach of these Terms, including the Content Standards, and the breach is not capable of remedy or is not remedied within a reasonable period after notice;

(b) your conduct creates a risk of harm to other users, to us or to a third party, including fraud, abuse, manipulation of quality or matching mechanisms, or the use of multiple Accounts;

(c) we are required to do so by law or by an order of a competent authority;

(d) you fail to pay an amount due and do not remedy the failure after notice;

(e) our payment provider terminates or refuses your account with it;

(f) you fail to provide information we are legally obliged to collect, including the tax information described in Section 44;

(g) your Account has been inactive for more than 24 months and you do not respond to a notice sent to your registered email address; before terminating for inactivity we will pay out any Wallet balance in accordance with Section 40.11.

17.3 Proportionality and notice. Suspension and termination are used proportionately, having regard to the seriousness of the breach, whether it is repeated, the harm caused and the fundamental rights of everyone concerned. Except where the breach is serious, unlawful or creates an immediate risk, we will warn you first and give you an opportunity to remedy. Every suspension and termination is accompanied by the statement of reasons required by Section 11.5 and may be challenged under Section 11.6.

17.4 What survives. Termination does not affect: any right or liability accrued before it; Sections 8.4 and 45 (confidentiality); Section 10 (intellectual property and licences, to the extent stated); Section 12 (personal data); Sections 18, 19 and 20 (disclaimers, liability, indemnity); Section 22 (governing law and disputes); and any other provision which by its nature is intended to survive.

17.5 Effect on running Experiments. If an Ideator Account is terminated for breach, Experiments in progress may be terminated. Validators who have already submitted a Response to a terminated Experiment are paid in full for it. Amounts committed to Responses not yet collected are refunded to the Ideator, less any amount properly retained under Section 28 or set off under Section 13.6.

17.6 Effect on a Wallet balance. Termination of a Validator Account does not extinguish a Wallet balance genuinely earned. Section 40.11 governs what happens to it, including where the Account is terminated for fraud.

17.7 Deletion of data. After termination we delete or anonymise your personal data in accordance with the Privacy Policy and the retention periods in Section 12.6, save where we must retain it to comply with a legal obligation, to establish, exercise or defend a legal claim, or to prevent fraud and abuse. Export your data before you close your Account — Section 35 tells you how.

18. Disclaimers of warranties

18.1 The Service is provided as it is. To the maximum extent permitted by applicable law, and subject always to Section 18.6, the Service and all AI Output are provided "as is" and "as available", without warranty of any kind, whether express, implied or statutory. We disclaim, to that extent, all implied warranties, including any warranty of merchantability, satisfactory quality, fitness for a particular purpose, title, accuracy, and non-infringement.

18.2 In particular, we do not warrant that:

(a) the Service will be uninterrupted, timely, secure or error-free;

(b) any defect will be corrected;

(c) the Service or its infrastructure is free of harmful components;

(d) any Experiment will attract any particular number of Responses, or any Response at all, within any particular time;

(e) any Response is truthful, considered, competent, representative or free from bias;

(f) any Report, verdict, grade, score, insight, recommendation or reasoning trace is accurate, complete, reliable, reproducible or fit for any decision you may take;

(g) the results of an Experiment predict, correlate with or bear any relationship to actual market behaviour, customer demand or commercial outcome;

(h) any Panel is statistically representative of any population or segment;

(i) any AI Output is original, protectable, or free from resemblance to output generated for another user.

18.3 No professional advice. Nothing produced by or through the Service is legal, financial, investment, tax, accounting, medical, psychological or regulatory advice. We are not a law firm, an investment adviser, a market research institute, an auditor or a certification body. Where a decision matters, take independent professional advice.

18.4 No guarantee of outcome for Validators. We do not guarantee that any Validator will receive any Offer, any particular number of Offers, any particular level of earnings, or any Offer at all. The volume of Offers depends on the Experiments that Ideators choose to run, on the profiles they target and on the number of Validators available. The Service is not a source of guaranteed or regular income and must not be relied on as one.

18.5 Third-party services. We do not warrant, endorse or assume responsibility for any third-party service, product, content or website that you access through or in connection with the Service, including app stores, identity providers, payment providers, artificial intelligence providers and integrations. Your dealings with them are between you and them.

18.6 Rights that cannot be excluded. Nothing in these Terms excludes or limits any warranty, guarantee, right or remedy that cannot be excluded or limited under applicable law. If you are a consumer, you have statutory rights in respect of digital services that are not in conformity with the contract, including the right to have the service brought into conformity, to a proportionate price reduction, or to terminate the contract. Those rights are unaffected by this Section. For the period fixed by law, the burden of proving conformity lies with us.

19. Limitation of liability

19.1 What is never limited. Nothing in these Terms excludes or limits our liability for: death or personal injury caused by our negligence; fraud or fraudulent misrepresentation; intent or gross negligence; any liability under mandatory product liability law; any liability that cannot lawfully be excluded or limited; and, where you are a consumer, any liability under mandatory consumer protection law.

19.2 Excluded categories of loss. Subject to Section 19.1, and to the maximum extent permitted by law, we are not liable for: loss of profit, revenue, business, contracts or anticipated savings; loss of goodwill or reputation; loss of opportunity; wasted expenditure, including expenditure incurred in reliance on a Report; loss, corruption or unavailability of data, except to the extent caused by our failure to take the security measures required by law; or any indirect or consequential loss, in each case however arising and whether or not foreseeable.

19.3 Decisions taken on the strength of a Report. Without limiting Section 19.2, we are not liable for any loss arising from a decision to build, not to build, to launch, not to launch, to invest, not to invest, to price, to pivot or to abandon anything, taken wholly or partly on the basis of a Report, a verdict, a grade, a score or any other AI Output. You accept that such output is an input to your judgement and not a substitute for it.

19.4 Financial cap. Subject to Section 19.1, our total aggregate liability arising out of or in connection with these Terms and the Service, whether in contract, tort (including negligence), breach of statutory duty, restitution or otherwise, in any period of twelve consecutive months, is limited to the greater of:

(a) the total amount you paid to us, or that we paid to you, under these Terms in the twelve months preceding the event giving rise to the liability; and

(b) one hundred euro (EUR 100).

19.5 Per-Experiment cap. Without prejudice to Section 19.4, our liability in respect of any single Experiment does not exceed the amount you paid for that Experiment.

19.6 Your liability to us. You are liable to us for loss we suffer as a result of your breach of these Terms, in accordance with general law. Where you are a consumer, your liability is limited to loss that was foreseeable at the time the contract was made.

19.7 Allocation of risk. You acknowledge that the price of the Service reflects the allocation of risk in this Section, and that we would not provide the Service on these commercial terms without it.

19.8 Time limit for claims. To the extent permitted by law, and except where you are a consumer, any claim arising out of or in connection with these Terms must be brought within one year of the date on which you became aware, or ought reasonably to have become aware, of the facts giving rise to it.

20. Indemnity

20.1 Your indemnity. Except where you are a consumer, you will indemnify us, our affiliates and our respective officers, employees and agents against all liabilities, damages, losses, costs and expenses, including reasonable legal fees, arising out of or in connection with:

(a) your User Content, including any claim that it infringes the rights of a third party or breaches an obligation of confidence;

(b) your breach of these Terms, of the Content Standards or of any applicable law;

(c) your use of any AI Output, including any claim arising from its publication, distribution or commercial use, and any failure by you to comply with an obligation that attaches to you as a deployer under Section 9.6;

(d) any personal data you introduce into the Service without a lawful basis;

(e) any claim brought by a person to whom you made the Service available in breach of Section 8.3(g).

20.2 Consumers. Section 20.1 does not apply to you if you are a consumer. Your liability to us is then determined by general law, as limited by Section 19.6.

20.3 Conduct of claims. We will notify you of any claim to which the indemnity applies, will not settle it without your consent (not to be unreasonably withheld), and will give you reasonable assistance at your expense. We may take control of the defence of a claim where our own reputation or intellectual property is at stake.

21. Changes to these Terms

21.1 We may change these Terms. We may amend these Terms to reflect a change in the Service, in our business model, in our providers, in applicable law, in the interpretation of applicable law by a court or authority, or in security or operational requirements, or to correct an error or an ambiguity.

21.2 Notice. We will publish the amended Terms with a new version number and effective date. Where the amendment is material — in particular where it changes prices, payment terms, the scope of the licences you grant, the limitation of liability, the governing law or the dispute mechanism — we will notify you at least 30 days before it takes effect, by email to your registered address and by a notice in the Service.

21.3 Your choice. If you do not accept a material amendment, you may terminate your Account, free of charge, at any time before it takes effect, and we will refund the unused portion of any amount you have paid in advance. Continuing to use the Service after the effective date constitutes acceptance. Where the law requires your express acceptance, we will ask for it and the amendment will not apply to you until you give it.

21.4 Immediate changes. Amendments required by law, by an order of an authority, or to address a security risk or an imminent harm may take effect immediately, with notice given as soon as reasonably possible.

21.5 Running Experiments. An amendment does not change the price, the fee or the payout of an Experiment that has already been paid for or launched. Those are fixed at the moment of payment or launch.

21.6 Version history. We keep every superseded version of these Terms. A copy of any previous version, and of the version you accepted, is available on request from gabri@validating.studio.

22. Governing law, jurisdiction and disputes

22.1 Governing law. These Terms, and any non-contractual obligation arising out of or in connection with them, are governed by the law of the Netherlands, excluding its conflict-of-law rules and excluding the United Nations Convention on Contracts for the International Sale of Goods.

22.2 Jurisdiction. The courts of Amsterdam, the Netherlands have exclusive jurisdiction over any dispute arising out of or in connection with these Terms or the Service.

22.3 Talk to us first. Before commencing proceedings, please contact us at gabri@validating.studio with a description of the dispute and the outcome you seek. We will respond within 30 days. Most disputes are resolved at this stage, and this step is a condition precedent to proceedings save where urgent relief is required or where a limitation period would otherwise expire.

22.4 Complaint before dispute. Where the dispute concerns a measure taken against your Account or your content, you must first use the internal complaint mechanism in Section 11.6, unless doing so would be manifestly futile.

22.5 Consumers. If you are a consumer, nothing in Sections 22.1 and 22.2 deprives you of the protection of the mandatory provisions of the law of the country in which you are habitually resident, and you may bring proceedings before the courts of that country. We may bring proceedings against you only before the courts of the country in which you are habitually resident.

22.6 Out-of-court dispute settlement. If you are a consumer and you are not satisfied with the outcome of our complaint handling, you may refer the dispute to a certified alternative dispute resolution body. PHASE 2 B.V. is not currently affiliated to any certified alternative dispute resolution body. You remain free to refer the dispute to a body competent for it in your country of residence, and we will engage with such a body in good faith. The European Commission's Online Dispute Resolution platform ceased operating on 20 July 2025 and is no longer available.

22.7 No class actions. To the extent permitted by law, and except where you are a consumer or where a collective action is provided for by mandatory law, disputes are resolved individually and not on a class, collective or representative basis.

23. General provisions

23.1 Entire agreement. These Terms, together with the documents incorporated by reference in Section 3, constitute the entire agreement between you and us in relation to the Service, and supersede all prior communications, representations and understandings on that subject. Nothing in this clause limits liability for fraudulent misrepresentation, or, where you are a consumer, the effect of any pre-contractual information we were required to give you, which forms part of the contract.

23.2 No waiver. A failure or delay in exercising a right is not a waiver of it, and a single or partial exercise does not preclude any further exercise.

23.3 Severability. If a provision of these Terms is held invalid, illegal or unenforceable, it is severed to the minimum extent necessary and the remainder continues in force. The parties will replace the severed provision with a valid one that comes closest to its commercial purpose.

23.4 Assignment. You may not assign or transfer these Terms, or any right or obligation under them, without our prior written consent. Your Account is personal to you. We may assign or transfer these Terms to an affiliate, or in connection with a merger, acquisition, reorganisation or sale of assets, on notice to you; where you are a consumer and the assignment reduces your rights, you may terminate free of charge.

23.5 No third-party rights. Except as stated in Section 16.6 in respect of Apple, and except for our affiliates and personnel in respect of Sections 19 and 20, no person other than you and us has any right to enforce these Terms.

23.6 Force majeure. Neither party is liable for a failure or delay in performance caused by an event beyond its reasonable control, including natural disaster, epidemic, war, civil unrest, act of terrorism, industrial action, failure of a public telecommunications network or of a utility, failure of a third-party provider, cyber-attack, embargo, sanction or act of government. The affected party will notify the other and will use reasonable efforts to mitigate. Payment obligations already accrued are not suspended.

23.7 Notices. We give you notice by email to the address registered on your Account, by notice within the Service, or by push notification. You give us notice by email to gabri@validating.studio. A notice sent by email is deemed received on the day it is sent unless the sender receives a delivery failure. Keep your email address up to date; you are responsible for notices sent to an address you have abandoned.

23.8 Language. These Terms are drawn up in English. Any translation is provided for convenience only, and in the event of a discrepancy the English version prevails, save where mandatory law in your country requires otherwise.

23.9 Independent parties. Nothing in these Terms creates a partnership, joint venture, agency, franchise or employment relationship between you and us.

23.10 Interpretation against the drafter. These Terms are not to be construed against us merely because we drafted them; but where a term is ambiguous and you are a consumer, the interpretation most favourable to you prevails.

23.11 Unfair terms in business contracts. Where these Terms are imposed unilaterally on a business user, any term relating to access to and use of data is to be read, and if necessary read down, so as not to be unfair within the meaning of Article 13 of Regulation (EU) 2023/2854.

PART B — ADDITIONAL TERMS FOR IDEATORS

This Part applies to you if you use the Service to create, fund or run Experiments. Part 0 and Part A apply to you as well.

24. Creating and launching an Experiment

24.1 The flow. You describe your idea; the Service helps you formulate a Hypothesis; you choose a Panel type; you define who you want to hear from; you build the Experiment out of question and task blocks; you set the acceptance criteria that will determine the verdict; you set a budget and pay. The Experiment is then reviewed under Section 11.1 and, if approved, launched.

24.2 The Hypothesis and the acceptance criteria are yours. The Service drafts them; you decide them. You may edit every question, every task and every acceptance criterion before you launch. Once launched, the Experiment is fixed: you cannot change the questions or the criteria, because doing so would invalidate the Responses already collected.

24.3 The verdict is mechanical. The verdict of an Experiment is the proportion of your make-or-break acceptance criteria that were met by the Responses actually collected, expressed as a grade and mapped to a verdict. If you set criteria that are easy to meet, you will get a positive verdict. If you set criteria that are impossible, you will not. The verdict measures your Experiment against your own criteria. It does not measure your idea against the world. Section 9.3 applies.

24.4 Materials you upload. You may attach files to an Experiment within the size and format limits shown in the interface. Sections 8, 10.3 and 10.4 apply to everything you attach. Files are scanned, but no scanning is perfect, and you are responsible for what you upload.

24.5 Review before launch. Every Experiment submitted to a Human Panel is reviewed before distribution, under Section 11.1. Approval is not an endorsement of your idea, a legal review of your content, or a warranty that the Experiment is well designed. It confirms only that the Experiment is not obviously incompatible with the Content Standards and is technically presentable.

24.6 Targeting. You may target an Experiment using anonymised professional and demographic criteria. You never receive the identity of a Validator, nor the raw value of any sensitive attribute (Section 12.4). Targeting must not be used to exclude or include persons on a basis that is unlawfully discriminatory, nor to identify a specific individual. We may refuse targeting that is so narrow that it would identify a person.

24.7 Panel type and its consequences. Choosing an Agentic Panel means no human being will evaluate your idea. Choosing a Human Panel means no machine perspective is collected. Choosing a Hybrid Panel produces two independent verdicts and a description of where they diverge; the divergence is an observation, not a third verdict, and neither side is presented as the correct one.

24.8 Iterating. You may create further Experiments to test a revised Hypothesis. Each Experiment is a separate purchase.

25. Prices, fees and taxes

25.1 Where the price lives. The price of an Experiment is calculated and displayed in full, itemised, before you pay. The itemisation shows the number of Responses, the amount attributable to Validator compensation, our platform fee, and any option you have selected. The total shown at the moment of purchase is the price of the contract. The Fee Schedule referred to in Section 3 sets out the prices, options, minimums and limits in force from time to time.

25.2 Options. Options such as accelerated distribution or restriction to higher-rated Validators are priced separately and shown as separate line items.

25.3 Minimum budget. A minimum budget applies to every Experiment and is stated in the Fee Schedule and enforced in the interface.

25.4 How Validators are paid, and what we keep. Part of what you pay is used to compensate Validators; the remainder is our platform fee, which covers the operation of the Service, the artificial intelligence compute, the payment processing and our margin. The split is shown to you in the receipt. The amount a given Validator receives is the amount stated in the Offer that Validator accepted, which is determined per Offer by our allocation mechanism and may differ between Validators for the same Experiment. You are never charged more than the total you approved.

25.5 Agentic Panels. For an Agentic Panel there is no Validator compensation. You pay for compute and our fee, priced per agent run.

25.6 Price changes. We may change prices for the future. A change never affects an Experiment already paid for, or a subscription period already paid for. Where a change affects a recurring subscription, Section 30.5 applies.

25.7 Taxes. Prices are stated inclusive or exclusive of value added tax as indicated in the interface; the total shown before payment is always the total payable, inclusive of all taxes and charges. Where you are a business established in a Member State other than ours and you provide a valid VAT identification number, the reverse charge may apply and the interface will reflect it. Where you are established outside the European Union, tax is applied according to the rules of your country of establishment or consumption. You are responsible for the accuracy of the tax information you give us, and for any tax that arises in your own hands as a result of your use of the Service.

25.8 Invoices. We issue an invoice for every payment, available in your account and sent to your registered email address. It is your responsibility to enter correct billing details before you pay; we will correct an invoice on request where the law allows.

25.9 No hidden charges. We do not charge anything that was not shown to you before you paid, other than a charge that arises from your own act — such as a chargeback fee under Section 13.4, or a tax you have caused to arise by giving incorrect information.

26. Payment and when the contract is concluded

26.1 The order. You place an order by completing the budget step and confirming payment. The confirmation button is labelled so as to make clear that the order carries an obligation to pay.

26.2 Authorisation and charge. On confirmation, a payment intent is created with our payment provider. The contract for the Experiment is concluded at that moment. Where the Experiment is subsequently rejected at review, Section 28.4 applies and you are refunded in full.

26.3 Failed payment. If payment fails, no Experiment is created and no contract is concluded.

26.4 Confirmation on a durable medium. We send you, by email, confirmation of the contract, containing the essential terms, the price paid, the itemisation, the confirmation of any statement you made under Section 34, and a copy of, or a link to, these Terms.

26.5 Credits. Where you hold Credits, they are applied against the total before your payment method is charged, unless the interface says otherwise.

27. What we guarantee, and what we do not

27.1 What we commit to. We commit to: distributing your Experiment to Validators matched to the criteria you selected, or executing the number of agent runs you purchased; collecting Responses in accordance with the Experiment you designed; generating a Report; making the results available to you; and charging you no more than the total you approved.

27.2 What we do not commit to — the number of Responses. An Experiment closes when the target number of Responses is reached, when the budget is exhausted, or when the pool of eligible Validators is exhausted, whichever comes first. We do not guarantee that any particular number of Responses will be collected. Where fewer Responses are collected than you purchased, Section 28.6 governs what happens to the unspent budget.

27.3 What we do not commit to — timing. The indicative turnaround for a Human Panel is 48 to 72 hours from launch, and for an Agentic Panel a matter of minutes. These are indications, not deadlines. Turnaround depends on the availability of Validators matching your criteria, on the narrowness of your targeting and on the size of the pool. Where accelerated distribution is purchased, it improves priority in the distribution queue; it does not create a deadline.

27.4 What we do not commit to — outcome. We do not commit to any verdict, grade, score, level of sentiment, quality of insight or usefulness of recommendation. Section 18 applies in full.

27.5 Failed or degraded runs. If an Agentic Panel fails to execute, or executes only in part, we will re-run it or refund the affected portion, at our option. If the artificial intelligence generation of a Report fails, we present the underlying structured data instead; the Report is never withheld for that reason, and no refund is due where the data and the verdict are delivered.

27.6 Quality of Responses. Responses come from independent individuals. Some will be more considered than others. We operate quality controls, but we do not warrant the quality of any individual Response. If a Response is manifestly non-responsive, unintelligible or fraudulent, tell us through the reporting control in the results view; where we agree, we will exclude it and either replace it or credit you the amount attributable to it.

28. Cancellation, refunds and unspent budget

28.1 Before launch. You may delete a draft Experiment at any time before payment, at no cost.

28.2 After payment, before approval. You may cancel an Experiment after payment and before it has been approved and distributed, and receive a full refund.

28.3 After launch. Once an Experiment has been distributed you may stop it at any time. Responses already submitted, and Offers already accepted and in the course of completion, are paid for in full and are delivered to you. The remaining unspent budget is refunded to your original payment method, or, at your choice, credited to your account.

28.4 Rejection at review. If we reject an Experiment at review under Section 11.1, we refund the full amount paid, and we tell you why. You may amend and resubmit.

28.5 Termination by us. If we terminate a running Experiment because it breaches these Terms, Responses already submitted are paid for and are delivered to you where lawful, and the unspent budget is refunded, unless the breach was fraudulent, in which case we may retain the amount required to compensate the Validators affected and to cover our reasonable costs.

28.6 Under-delivery. Where an Experiment closes with fewer Responses than you purchased, the portion of the budget attributable to the Responses not collected is refunded automatically to your original payment method, or credited to your account at your choice, within 14 days of the Experiment closing. You are never charged for a Response you did not receive.

28.7 Excluded Responses. Where a Response is excluded under Section 27.6, the amount attributable to it is refunded or credited on the same basis.

28.8 How refunds are made. Refunds are made to the original payment method, in the original currency, without undue delay and in any event within 14 days of the event giving rise to them. Credits used in the original payment are refunded as Credits. We do not charge a fee to make a refund.

28.9 Beyond these cases. Outside the cases described in this Section, in Section 27.5, in Section 30.6 and in Section 34, payments for Experiments are not refundable, because the Service has been delivered: Validators have been compensated for work genuinely done and compute has been consumed. This does not affect your statutory rights as a consumer.

28.10 Disputed quality. If you believe an Experiment as a whole was not delivered in conformity with the contract, use the complaint mechanism in Section 11.6 and Section 22.3 before raising a chargeback.

29. Credits and referrals

29.1 What Credits are. Credits are a promotional balance that may be applied against the price of an Experiment. Credits are not money, are not redeemable for money, cannot be transferred, sold or gifted, and confer no proprietary right.

29.2 Referrals. Where we operate a referral programme, its rules — who qualifies, what triggers the credit, and how much — are published in the Service, in the place where the programme is offered, and may be changed for the future. Credits are typically awarded only once the person you referred has completed a qualifying action, not merely signed up.

29.3 Fair use. Credits obtained through self-referral, multiple Accounts, false identities, automated sign-ups or any other manipulation are void, and the underlying conduct is a breach of Section 8.3.

29.4 Expiry and withdrawal. We may set an expiry period for Credits awarded in the future; any expiry will be stated when the Credits are awarded. We may discontinue the referral programme at any time, on notice, without affecting Credits already awarded.

29.5 Order of application. Credits are applied before your payment method is charged. Where Credits exceed the price, the excess remains as Credits.

30. Subscriptions

30.1 What a subscription is. We may offer a paid subscription giving access to additional features. The features included, the price and the billing period are described at the point of purchase and in the Fee Schedule.

30.2 Term and renewal. Unless stated otherwise, a subscription runs for the period purchased and renews automatically for successive periods of the same length, at the price then in force, until cancelled. We send a renewal reminder at least 14 days before each renewal, stating the date and the amount.

30.3 Cancellation. You may cancel a subscription at any time, with effect from the end of the current period. You keep the subscription features until then. Cancellation is available in your account settings, or through the app store where the subscription was purchased there.

30.4 Failed renewal payment. If a renewal payment fails, we will retry and notify you. You have a grace period of 7 days to update your payment method, during which the subscription features remain active. After that, the subscription lapses and the features are removed.

30.5 Price changes on renewal. We may change the subscription price for future periods. We will notify you at least 30 days before the renewal to which the new price applies. If you do not accept the new price, cancel before the renewal date; otherwise the renewal takes effect at the new price.

30.6 Refunds. Subscriptions are paid in advance for the whole period and are not refundable pro rata on cancellation, except: where you exercise a right of withdrawal under Section 34; where we materially reduce the subscription features under Section 14.5; where we terminate your Account other than for your breach; or where a mandatory rule of consumer law provides otherwise.

30.7 Free trials and promotions. Where we offer a free trial or a promotional price, the conditions are stated at the point of offer, including what happens at the end of the trial. If a trial converts into a paid subscription, we will tell you before it does, and how to prevent it.

30.8 Store-based subscriptions. Where the subscription was purchased through an app store, Section 16.8 governs billing, renewal, cancellation and refunds, and you must exercise those rights through the store.

30.9 Effect of losing a subscription. Data created while you had a subscription remains yours. Features that require a subscription — such as advanced analytics, custom branding, exports or API access — cease to be available when it ends. Export what you need first (Section 35).

31. API and developer access

31.1 Availability. The API is made available to subscribers as described in the Fee Schedule and in the developer documentation published at api.validating.studio/v1/docs when the API is released, which forms part of these Terms in respect of your use of the API.

31.2 Keys. API keys are personal to the Account that created them, are not transferable, and are shown to you once. You are responsible for keeping them secret and for everything done with them. Notify us immediately if a key is exposed; we will revoke it.

31.3 Rate limits. Requests are subject to the rate limits published in the documentation. Exceeding a limit results in a temporary refusal of service. Deliberate circumvention of a rate limit is a breach of Section 8.3(c).

31.4 Permitted use. You may use the API to integrate the Service with your own systems. You may not use it to: build a product that substitutes for or competes with the Service; extract, mirror or resell data at scale; train, evaluate or benchmark a machine-learning model; or provide access to a third party outside your Workspace.

31.5 Webhooks. Where you register a webhook endpoint, you are responsible for its security, for verifying the signature we send, and for handling duplicates and retries idempotently. We retain delivery logs for a limited period stated in the documentation.

31.6 Versioning and deprecation. We support a released major version of the API for at least twelve months after the release of its successor. We announce deprecations in the documentation and by email to the Accounts using the affected version. Breaking changes are made in a new major version, except where a change is required for security or legality.

31.7 Suspension. We may suspend or revoke API access immediately where its use threatens the security, integrity or availability of the Service, breaches these Terms, or exceeds limits persistently after notice.

32. Third-party integrations

32.1 Connecting a third-party service. You may connect certain third-party services, such as a messaging or documentation tool. Doing so authorises us to exchange with that service the data necessary for the integration to work, within the scopes you approve.

32.2 Their terms, their responsibility. The third-party service is provided by the third party under its own terms and privacy policy. We do not control it, do not warrant it, and are not responsible for it or for what it does with data you send to it.

32.3 Disconnecting. You may disconnect an integration at any time, in the Service or in the third-party service. On disconnection we stop sending data and delete the stored authorisation. Data already delivered to the third party remains subject to that party's terms; ask them to delete it.

32.4 Withdrawal of an integration. We may discontinue an integration where the third party changes or ends its interface, or where continuing it would create a security or legal risk.

33. Workspaces and team access

33.1 What a Workspace is. A Workspace allows several people to access the same Experiments under a single subscription, with distinct roles: an owner with full access including billing; administrators who may manage members and Experiments; and viewers with read-only access.

33.2 The owner is responsible. The Workspace owner is responsible for who is invited, for the conduct of every member in the Workspace, for the payment of the subscription, and for ensuring that every member accepts these Terms.

33.3 Ownership of content. Experiments and Reports created within a Workspace belong to the Workspace, not to the individual who created them. If a member leaves, or is removed, the Experiments they created remain in the Workspace and remain accessible to it. Members should not create in a Workspace anything they wish to retain personally.

33.4 Confidentiality within a Workspace. Every member has access to the Experiments in the Workspace and is bound by Section 8.4 in respect of them.

33.5 Removal and closure. The owner may remove a member at any time. If the Workspace subscription ends, the Workspace becomes read-only and, after the period stated in the Fee Schedule, its data is deleted. Export first (Section 35).

33.6 Limits. The maximum number of members is stated in the Fee Schedule.

34. Right of withdrawal (consumers only)

34.1 Who this Section applies to. This Section applies only if you are a consumer — a natural person acting for purposes outside your trade, business, craft or profession — and you are resident in the European Union or in another jurisdiction granting an equivalent right. We do not require you to declare that you are a professional, and a declaration to that effect does not deprive you of consumer rights if in fact you are a consumer.

34.2 The right. You have the right to withdraw from a contract concluded with us at a distance, without giving any reason, within 14 days of its conclusion.

34.3 How to withdraw. Inform us of your decision by an unequivocal statement — for example by email to hello@validating.studio, or by using the model form in Annex A. Sending the statement before the 14-day period expires is sufficient. We will acknowledge receipt without delay on a durable medium.

34.4 Effect. We will reimburse all payments received from you, without undue delay and in any event within 14 days of being informed of your withdrawal, using the same means of payment you used, at no cost to you — subject to Sections 34.5 and 34.6.

34.5 Immediate performance, and how you lose the right. Running an Experiment starts immediately, and Validators begin to be compensated as soon as it is distributed. If you want it to start before the 14-day period ends, we must ask you for two things, and we ask for them before you pay, in separate, non-pre-ticked boxes:

☐ I expressly request that the performance of the Service begin immediately, before the end of the 14-day withdrawal period.

☐ I acknowledge that I will lose my right of withdrawal once the Service has been fully performed.

Where the Service consists in the supply of digital content not supplied on a tangible medium — such as the download of a Report — a third acknowledgement is requested:

☐ I acknowledge that I will lose my right of withdrawal once the supply of that digital content has begun.

34.6 If you withdraw after performance has begun. If you gave the request and acknowledgements above and you withdraw before the Service has been fully performed, you must pay an amount proportionate to what has been supplied up to the moment you informed us of your withdrawal, calculated on the basis of the total price agreed. In practice this means the Responses already collected and the compute already consumed. If we did not obtain the request and acknowledgements above, you owe nothing at all, even if the Service was performed.

34.7 Subscriptions. A subscription may be withdrawn from within 14 days of purchase on the same basis. If you asked us to begin immediately, you pay for the part of the period already elapsed.

34.8 Confirmation. We confirm the request and the acknowledgements you gave under Section 34.5, together with the contract, on a durable medium, by email, before performance begins.

35. Export, portability and continuity

35.1 Export. You may export, at any time and at no cost, in a structured, commonly used and machine-readable format: your Experiments, your Hypotheses, your acceptance criteria, the aggregated Responses to your Experiments, your Reports and the associated metadata.

35.2 On termination. After termination of your Account or of a Workspace, we keep your data available for export for at least 30 days, unless we are required to delete it sooner or the termination was for fraud.

35.3 Switching. We will give you reasonable assistance, on request, to transfer your data to another provider or to your own infrastructure. We do not charge an exit fee.

35.4 No lock-in by format. We will not deliberately structure exports so as to make them unusable elsewhere.

PART C — ADDITIONAL TERMS FOR VALIDATORS

This Part applies to you if you use the Service to receive Offers and to be paid for completing Experiments. Part 0 and Part A apply to you as well.

36. Your status: independent, occasional, and free

36.1 What this relationship is. You provide Responses to Experiments as an independent person, on your own account, occasionally and entirely at your own discretion. These Terms do not create a contract of employment, a contract of service, an agency, a partnership or any other relationship of subordination between you and us. You are not our employee, worker, agent or representative, and you must not present yourself as one.

36.2 No obligation to accept work. You are free to accept or to refuse any Offer, at any time, for any reason or for no reason, without explanation. Refusing an Offer, or simply not responding to one, is not a breach of any obligation, is not recorded as a fault, and has no negative effect on your reliability score, on the priority with which you receive future Offers, on your eligibility for any category of Experiment, or on your access to the Service.

36.3 No minimum availability. There is no minimum number of Experiments you must complete, no minimum time you must be available, no shift, no rota, no required response rate and no obligation to log in. You may be inactive for as long as you wish, subject only to the inactivity provision in Section 17.2(g).

36.4 No exclusivity. You may provide the same or similar services on any competing platform, and to your own clients, at the same time. We will never ask you not to.

36.5 You organise your own work. You decide whether, when, where, on what device and for how long you use the Service. We tell you what result we need — a considered answer to the questions in the Experiment — and not how, when or where to produce it.

36.6 No disciplinary power. We do not exercise disciplinary authority over you. The measures described in Section 46 are contractual measures concerning the conformity of the service you provide with what was agreed and with these Terms. They are not sanctions on your person or your conduct as an individual.

36.7 Your own equipment and costs. You use your own device and your own connectivity. We do not reimburse costs unless we expressly agree to in writing.

36.8 What determines your remuneration. The amount payable for a given Offer is stated in that Offer, in full, before you accept it. You are never asked to work for an amount you have not seen and accepted.

36.9 Reclassification. If, notwithstanding this Section, a court or authority in your country determines that our relationship is one of employment or of another protected status, that determination applies from the date it specifies and does not retroactively change the amounts already paid to you, which will be treated as gross amounts on account of any sum found to be due.

36.10 Mandatory law prevails. Nothing in this Section deprives you of any right conferred by a mandatory rule of the law applicable to your situation, including any rule implementing Directive (EU) 2024/2831 on improving working conditions in platform work. Where such a rule confers a right or a protection that these Terms do not, that rule applies.

37. Becoming a Validator

37.1 Onboarding. To become a Validator you complete a structured onboarding in the mobile application, in which you provide your name and email address, your country of residence, your age range, and information about your profession, your field, your skills, your interests and your availability. You give a separate, explicit consent to the processing of that data, which is required before you can proceed.

37.2 Why we ask. The information you give determines which Experiments you are matched to. It is the single most important factor in whether you receive Offers, and in the value of the Service to Ideators. Accurate answers are the basis of the whole arrangement. Section 6.3 applies.

37.3 Supporting materials. We may ask you to substantiate your professional profile, for example with a curriculum vitae or a link to a professional profile page. Where accepted alternatives exist, they are listed in the application at the point of the request. The materials are assessed with the assistance of an automated quality check, described in Section 42.

37.4 Approval. We review applications and may approve, refuse or defer them. Where we refuse, we tell you the reason within the categories we apply, and you may ask for a human review under Section 9.9 and may re-apply once the reason for refusal no longer applies. Approval is not a promise of any volume of Offers or of any level of earnings; Section 18.4 applies.

37.5 Keeping your profile current. You must update your profile when your circumstances change — a new role, a new country of residence, a change of field. A profile that no longer reflects reality degrades the Service for everyone and may lead to the measures in Section 46.

37.6 Waitlist. We may operate a waitlist where the number of applications exceeds the number of Experiments available. Being on a waitlist creates no entitlement.

38. Offers, the acceptance window, and completion

38.1 How you receive work. When an Experiment matches your profile you receive an Offer, normally by push notification. The Offer states what the Experiment is about at a general level, the estimated time to complete it, and the exact amount you will receive if you complete it.

38.2 The acceptance window. An Offer is valid for a limited period, currently two minutes, after which it is automatically re-allocated to another Validator. This window is the technical validity of the Offer. It exists so that Ideators receive Responses within a usable time and so that the Offer can be passed to someone else. It is not a working-time obligation. Letting an Offer expire is not a refusal, is not recorded against you, and has no consequence of any kind (Section 36.2).

38.3 Improved Offers. If an Experiment does not find enough Validators at first, you may receive a further Offer for the same Experiment at a higher amount. If you do, the application will say so expressly. You are free to accept or decline that one too, on the same terms.

38.4 Completing an Experiment. Once you accept an Offer you may complete the Experiment at your own pace, within the time indicated. If you abandon it, no payment is due for it, and nothing else follows.

38.5 What we ask of a Response. A Response must be your own, genuine, considered opinion, formed after actually engaging with the material. You must not: use an automated tool to generate answers; answer at random or without reading; copy answers from another source; complete the same Experiment more than once; or answer in a language you do not understand.

38.6 Honesty is the product. Ideators buy honest reactions. A negative, sceptical or critical Response is exactly as valuable to us as a positive one, and is never penalised. Nothing in the Service rewards you for liking an idea.

38.7 Content you may see. Experiments contain unpublished ideas, mock-ups and prototypes. Section 45 governs what you may do with them. If an Experiment contains content you consider unlawful, offensive or distressing, exit it and report it using the in-application control; you will not be penalised for doing so, and you may still be paid for a Response already submitted where it is fair to do so.

38.8 Special categories of data. An Experiment must not ask you for information about your health, your beliefs, your political opinions, your sex life or other special categories of personal data unless the Experiment has been expressly approved for that purpose and asks for your specific, explicit and revocable consent. You may always decline. Declining has no consequence. Report any Experiment that asks for such data without asking for consent.

39. What you are paid

39.1 The amount. You receive exactly the amount shown in the Offer you accepted. There is no percentage for you to compute, no deduction applied afterwards, and no adjustment after the fact, save where Section 43 applies.

39.2 How the amount is set. The amount is determined per Offer by our allocation mechanism, having regard to the Experiment, the profile requested, and the state of the pool of eligible Validators at that moment. It may therefore differ between Experiments, and between Validators for the same Experiment. Whatever it is, it is shown to you in full before you accept.

39.3 When it is earned. The amount is earned when you submit a complete Response to an Experiment you accepted. It is credited to your Wallet when the Experiment completes.

39.4 No other remuneration. Unless we agree otherwise in writing, the amount in the Offer is the whole of your remuneration. There is no bonus, no tip, no commission and no expense reimbursement.

39.5 Programmes and benefits. If we introduce a paid programme with additional features for Validators, its terms will be published at the point of purchase. No benefit that is not currently offered in the application is promised, committed or forthcoming, whatever you may have read elsewhere.

39.6 Currency. Amounts are stated and paid in euro (EUR) unless the application states otherwise for your country. Section 13.5 applies to conversion.

40. Wallet and payouts

40.1 The Wallet. Amounts earned are recorded in your Wallet, which shows your current balance, your lifetime earnings, the date of your last payout, and a history of transactions in which each Experiment appears under an anonymised label.

40.2 The Wallet is a record, not a bank account. The Wallet is a record of what we owe you. It is not a payment account, it bears no interest, it is not covered by any deposit guarantee scheme, and no funds are held by us on your behalf as a payment service.

40.3 Cashing out. You request a payout in the application. A minimum balance applies, currently twenty euro (EUR 20), and is stated in the Fee Schedule. Below that amount the payout control is disabled; the balance is not lost and remains available until you reach the threshold.

40.4 Bank details. To receive a payout you must complete the onboarding of our payment provider and connect a bank account in your own name. We cannot pay a bank account belonging to someone else. You are responsible for the accuracy of the details you provide; a payment sent to a wrong account you supplied cannot always be recovered.

40.5 Timing. Payouts are initiated promptly once requested and once the conditions in this Section are met. Transfers typically arrive within two business days, but this is an indication and not a guarantee; the actual time depends on your bank, on your country and on our payment provider.

40.6 Fees. Any fee charged by our payment provider on a payout, and any currency conversion fee, is disclosed in the Fee Schedule and, where applicable, at the moment you request the payout. We do not charge our own fee for making a payout, and we do not apply an inactivity or dormancy fee to a Wallet balance.

40.7 Failed payouts. If a payout fails, we retry and notify you. Persistent failure normally indicates a problem with the bank details or with the payment provider's verification, and we will tell you what to correct.

40.8 Eligibility for payouts. Payouts require identity verification (Section 41), a supported country of residence, and the tax information described in Section 44. Where any of these is missing, the balance remains in your Wallet until it is supplied.

40.9 Taxes. Amounts paid to you are gross. Section 44 applies.

40.10 When a payment may be suspended. We may suspend a payout, or place a Wallet balance on hold, only where: we have a reasonable and specific suspicion of fraud, of the use of multiple Accounts, or of automated or copied Responses; a legal obligation or an order of an authority requires it; our payment provider has blocked the transfer; or the tax information required by Section 44 has not been supplied after two reminders. Any suspension is notified to you with reasons under Section 11.5, is reviewed by a human being, and is resolved within 14 days of your request for review under Section 9.9. A suspension that is not confirmed on review is lifted immediately and the payout made.

40.11 Your balance if your Account ends. If you close your Account, or if we close it, the balance genuinely earned remains yours and will be paid out, if necessary below the minimum threshold, once identity and tax requirements are met. This applies even where the Account is closed for breach, except in respect of amounts earned by the fraudulent conduct itself, which we may withhold and set off under Section 13.6. Where a balance cannot be paid because you cannot be contacted or the payment provider cannot complete the transfer, we hold it and continue to attempt payment; it is not forfeited to us, and is dealt with in accordance with the law on unclaimed sums.

40.12 Deceased or incapacitated Validators. Where we are notified of the death or incapacity of a Validator, we will deal with the balance in accordance with applicable succession or protection law, on production of appropriate documentation.

41. Identity verification

41.1 Why. We verify the identity of Validators in order to pay them, to comply with our legal obligations, and to keep the pool free of duplicate and fraudulent Accounts.

41.2 When. Identity verification is not requested during onboarding. It is requested later, at the point where you have been matched to Experiments and verification is what stands between you and being paid.

41.3 How. Verification is carried out by Stripe Identity, using a government-issued document and, where required, a photograph of your face compared one-to-one with the document. We do not receive or store your identity document. The biometric comparison is performed by the provider for the sole purpose of confirming that the document belongs to you, and the resulting data is not used for any other purpose, is not used to identify you in any other context, and is not shared with Ideators.

41.4 Consent. The processing of biometric data for this purpose requires your explicit consent, which is requested at the moment of verification. If you do not consent you cannot be paid, but nothing else follows — your Account is not closed and your existing balance is not forfeited; it remains available if you later verify.

41.5 Failure or refusal. If verification fails, you may retry. If it fails repeatedly, or if the documents indicate that you are below the minimum age or that the Account is not yours, Section 5.4 and Section 46 apply.

42. Automated systems that affect you, and your rights

42.1 Full disclosure. The following automated systems are used in relation to Validators. This Section, together with the AI Transparency Notice, is the disclosure required by data protection law and by the rules on algorithmic management in platform work.

• Matching engine — What it does: Determines which Validators receive an Offer for a given Experiment and in what order · Main inputs: Your declared industry, sub-industry, job title, skills, interests, country, age range, availability; the targeting chosen by the Ideator; your reliability score; whether you are currently eligible · Main effect on you: How many Offers you receive, and for which Experiments
• Profile quality check — What it does: Assesses the substance and internal consistency of the professional materials you submit · Main inputs: The materials you upload or link · Main effect on you: Whether your application is approved, deferred or refused
• Response quality check — What it does: Assesses whether a Response engages with the Experiment · Main inputs: The content and structure of your answers, time taken · Main effect on you: Whether a Response is flagged for human review under Section 43
• Reliability score — What it does: Aggregates your history of completed Experiments into a single figure · Main inputs: Completion of accepted Offers, outcomes of quality checks, confirmed reports of misconduct · Main effect on you: Eligibility for Experiments restricted to higher-rated Validators, and therefore access to some better-paid work
• Fraud and abuse detection — What it does: Flags patterns indicating multiple Accounts, automated answers or manipulation · Main inputs: Device, account, behavioural and payment signals · Main effect on you: Whether your Account is reviewed, restricted or suspended

42.2 What does not feed these systems. Refusing an Offer, letting an Offer expire, abandoning an Experiment, giving negative or critical answers, reporting an Experiment, exercising any right under these Terms, or complaining, is never an input to any of the systems above and never counts against you.

42.3 What we never process. We do not process, in relation to Validators: your emotional or psychological state; your private conversations; data used to predict whether you would exercise collective rights or join a trade union; or data revealing racial or ethnic origin, migration status, political opinions, religious or philosophical convictions, disability, health, sex life or sexual orientation. We do not use biometric data except for the one-to-one verification described in Section 41.

42.4 Human oversight. The systems above are monitored by qualified staff who have the authority and the ability to override their output.

42.5 Human review of decisions that matter. Every decision that restricts, suspends or closes your Account, refuses or suspends a payment, or changes the terms on which you receive Offers, is reviewed by a qualified human being on your request. Write to gabri@validating.studio. We will give you written reasons, consider anything you submit, and complete the review within 14 days.

42.6 Correction and compensation. If the review shows that the decision was wrong, we reverse it immediately, restore what can be restored — including reinstating eligibility and releasing any suspended payment — and, where you have suffered quantifiable loss as a direct result, we will discuss appropriate compensation.

42.7 No retaliation. You will not be disadvantaged in any way for exercising a right under these Terms, for making a complaint, for reporting content, or for contacting a supervisory authority or a court.

42.8 Explanation. You may ask us, at any time, to explain how the systems above operate in relation to you, in plain language, including the main criteria used and their relative importance. Section 9.8(c) applies.

43. Quality, and what happens if a Response is challenged

43.1 The principle. A Response that you completed genuinely is paid for. We do not withhold payment because an Ideator dislikes the answer.

43.2 When a Response may be rejected. A Response may be rejected, and the corresponding amount not credited, only where it is: manifestly non-responsive, such as random, empty or unintelligible answers; produced by an automated tool; copied from another source or from another Response; submitted for an Experiment you were not eligible for; or a duplicate.

43.3 Procedure. Before a Response is rejected, the case is reviewed by a human being. If it is rejected, we tell you which Response, why, and how to contest it. You may contest it under Section 42.5 and we will answer within 14 days. If you are right, the amount is credited immediately.

43.4 Repeated rejection. Repeated rejection of Responses may lead to the measures in Section 46, following the same procedure and with the same rights of review.

43.5 Reliability score transparency. You may ask us at any time what your current reliability score is, what has affected it, and what would improve it. We will tell you.

44. Taxes and reporting

44.1 You are responsible for your own taxes. Amounts paid to you are paid gross. You are solely responsible for declaring them and for paying any income tax, social security contribution or other levy that arises in your country. We do not withhold tax and we give no tax advice. If you are unsure of your position, take advice; the amounts involved may nonetheless have to be declared.

44.2 Registration. In some countries, receiving payments of this kind on a regular basis may require you to register in a particular capacity. That is your responsibility to assess.

44.3 Information we must collect and report. Under Council Directive (EU) 2021/514 and its national implementations, and under equivalent rules in other jurisdictions, we may be required to collect, verify and report to the tax authorities information about Validators who receive payments, including: your name, address, date of birth, tax identification number and the Member State that issued it, the identifier of the financial account to which you are paid, and the total amount paid to you in each quarter, together with any fee withheld.

44.4 Your obligation to provide it. You must provide that information when asked and must tell us promptly if it changes. If you do not provide it after two reminders, and a period of at least 60 days has elapsed, we must suspend payouts and may close your Account.

44.5 A copy for you. Where we report information about you, we will make available to you a copy of the information reported, by the deadline set by the applicable rules.

44.6 Accuracy. You are responsible for the accuracy of the tax information you give us. We may rely on it.

45. Confidentiality

45.1 The obligation. Everything you see in an Experiment that is not public — the idea, the problem statement, the proposed solution, the pricing, the mock-ups, the prototypes, the images, the file attachments, the names of competitors listed and any commercial information — is confidential information of the Ideator, and you receive it in confidence.

45.2 What you must not do. You must not, during or after your use of the Service: disclose that information to any person; publish it, post it, describe it or discuss it in public, on social media, in a forum, in a chat group or to a journalist; photograph, screenshot, screen-record or otherwise copy it; use it for your own purposes or for the purposes of any other person, including to build, market or invest in a competing product; or attempt to identify or contact the Ideator.

45.3 Duration. This obligation applies from the moment you see the material and continues for as long as the material is not public through no act or omission of yours, and in any event survives the closure of your Account.

45.4 Exceptions. The obligation does not apply to information that is already public, that you already lawfully knew without an obligation of confidence, that you develop entirely independently, or that you are required to disclose by law or by an order of an authority — in which case, where lawful, tell us first.

45.5 Consequences. Breach of this Section is a material breach of these Terms. It may result in the immediate closure of your Account and in a claim for damages by us or, directly, by the Ideator whose information you disclosed, who may enforce this Section against you.

45.6 Reporting rather than disclosing. If an Experiment contains something you believe is unlawful or dangerous, do not publish it. Report it to us under Section 11.2 or, where appropriate, to a competent authority. Nothing in this Section prevents you from reporting a suspected breach of law to an authority or from exercising a right conferred by whistleblower protection law.

46. Measures, suspension and closure of a Validator Account

46.1 The measures available to us. Where you breach these Terms, we may, proportionately: warn you; require you to correct your profile; exclude a Response under Section 43; adjust your reliability score; reduce or suspend the Offers you receive; suspend payouts under Section 40.10; suspend your Account; or close it.

46.2 Grounds. The grounds are those in Section 17.2 and, specifically: false or materially inaccurate profile information; automated, copied or non-genuine Responses; multiple Accounts; acting as a Validator on an Experiment connected to you; breach of Section 45; abusive conduct towards our staff; and failure to supply the information required by Sections 41 and 44.

46.3 Procedure. Except where the breach is fraudulent, unlawful or creates an immediate risk, we warn you first and give you an opportunity to correct. Every measure is notified to you with a statement of reasons under Section 11.5, is reviewable by a human being under Section 42.5 within 14 days, and may be complained against under Section 11.6.

46.4 Transparency of restrictions. Where we reduce or suspend the Offers you receive, we will tell you that we have done so, and why. We will not leave you to guess.

46.5 Your balance. Section 40.11 governs what happens to a Wallet balance on suspension or closure.

46.6 Re-application. Where an Account has been closed other than for fraud or for conduct causing serious harm, you may ask us to re-open it or to allow a new application once the ground for closure no longer applies.

PART D — ANNEXES

Annex A — Model withdrawal form

Complete and return this form only if you wish to withdraw from the contract. See Section 34.

To: PHASE 2 B.V., De Boelelaan 1095a, Room 0A16, 1081 HV Amsterdam, the Netherlands — hello@validating.studio

I hereby give notice that I withdraw from my contract for the supply of the following service:

______________________________________________________________

Ordered on / received on: ______________________________________

Name of consumer: _____________________________________________

Address of consumer: __________________________________________

Order or invoice reference: ____________________________________

Signature of consumer (only if this form is notified on paper): ____________________

Date: _____________________

Annex B — Content Standards

These Standards form part of the Terms. They apply to everything submitted, uploaded, generated, requested or distributed through the Service, by Ideators and by Validators alike. Section 8 summarises them; this Annex is the operative list.

B.1 Content that is prohibited absolutely

The following is prohibited without exception, and its presence will result in immediate removal, immediate termination and, where applicable, a report to the competent authorities:

1. Child sexual abuse material, and any content that sexualises a person under 18 or presents them in a sexualised manner, whether real, drawn, animated or synthetically generated.
2. Non-consensual intimate images, including synthetic or manipulated images of that nature.
3. Content that incites, facilitates, instructs on or glorifies terrorism, mass violence, or violence against an identifiable person or group.
4. Content that incites hatred or discrimination against a person or group on the basis of race, ethnic origin, nationality, religion, disability, age, sex, sexual orientation, gender identity or any other protected characteristic.
5. Content facilitating human trafficking, forced labour, or the sexual exploitation of any person.
6. Instructions for the manufacture of weapons, explosives, or chemical, biological, radiological or nuclear agents.
7. Malicious code, phishing material, credential-harvesting material, or anything designed to compromise a device or a system.
8. Content that is unlawful under Union law or the law of any country in which it is accessible.

B.2 Content that is prohibited in this Service

The following is prohibited because it is incompatible with the purpose of the Service or with the interests of the people who use it:

1. Content designed to test, promote, market or develop a product, service or activity that is illegal, or the marketing of which is prohibited, in the market to which the Experiment relates — including unlicensed financial services, unlicensed gambling, unapproved medicines and medical devices, controlled substances, unlicensed weapons, and services that facilitate the evasion of law.
2. Content that infringes any intellectual property right, trade secret, image right or right of personality of a third party.
3. Content disclosed in breach of an obligation of confidence owed to a third party.
4. Personal data of a third party included without a lawful basis, and any special category of personal data outside the framework of Section 10.6.
5. Content that is gratuitously shocking, graphic, obscene or distressing, having regard to the fact that Validators receive Experiments without choosing their content in detail.
6. Content that impersonates a person, organisation or brand, or that falsely suggests an affiliation, endorsement or sponsorship.
7. Content presented as human-authored where it was generated by an artificial intelligence system and the presentation would mislead.
8. Content whose real purpose is to solicit, recruit, sell, advertise or generate leads rather than to test a hypothesis.
9. Content designed to elicit a political opinion, a religious belief or a vote, or to test political messaging, unless we have expressly approved it in advance in writing.
10. Content directed at, or designed to be answered by, persons under 18.
11. Content designed to identify a specific individual, or targeting so narrow that it would in practice do so.

B.3 Conduct that is prohibited

1. Creating or operating more than one Account of the same type; using another person's identity or credentials; allowing another person to use your Account.
2. Acting as a Validator on an Experiment created by you, by a person connected to you, or by an organisation in which you have an interest.
3. Coordinating with other users to influence the outcome of an Experiment, or offering or accepting any inducement to answer in a particular way.
4. Submitting Responses generated by an automated tool, copied from another source, or produced without genuinely engaging with the material.
5. Manipulating, gaming or attempting to reverse engineer any matching, quality, reliability or pricing mechanism.
6. Scraping, systematically extracting, mirroring or bulk-downloading content from the Service; accessing the Service by automated means outside the API; circumventing a rate limit or an access control.
7. Using the Service, or anything obtained through it, to train, fine-tune, evaluate or benchmark a machine-learning model, or to build a dataset for that purpose.
8. Reverse engineering, decompiling or disassembling any part of the Service, or attempting to extract prompts, model parameters, matching logic or scoring logic, except to the strict extent that such an act cannot lawfully be prohibited.
9. Removing, altering, obscuring or degrading any provenance marking, watermark, metadata or label identifying content as artificially generated or as originating from the Service.
10. Disclosing, copying, photographing, screenshotting, recording, publishing or exploiting the confidential content of an Experiment (Sections 8.4 and 45).
11. Attempting to identify, contact or approach a Validator, or an Ideator, outside the Service.
12. Probing, scanning or testing the security of the Service without our prior written authorisation, or interfering with its operation or with its use by others.
13. Abusive, threatening or harassing conduct towards our staff, or towards another user.
14. Using the Service in breach of a sanctions regime, or from a restricted territory (Section 5.7).
15. Using the Service for any practice prohibited by Article 5 of Regulation (EU) 2024/1689.

B.4 How these Standards are enforced

Enforcement follows Section 11: prior review of Experiments, notice and action, proportionate measures, a statement of reasons for every measure, and an internal complaint that a human being decides. Repeated or serious breach leads to termination. Where a breach appears to constitute a criminal offence involving a threat to life or safety, we inform the competent authorities.

PART E — AI TRANSPARENCY NOTICE

This Part is the AI Transparency Notice referred to in Section 3. It forms an integral part of these Terms and is published pursuant to Article 50 of Regulation (EU) 2024/1689 (the AI Act), Articles 13, 14, 15 and 22 of Regulation (EU) 2016/679 (the GDPR), and the transparency requirements applicable to algorithmic management.

E.1 In one paragraph

Large parts of Validating are operated by artificial intelligence. When you write your idea, an AI turns it into a hypothesis. When you build an experiment, an AI drafts the questions and the criteria. When your experiment finishes, an AI writes the report and computes the verdict. If you choose an agentic panel, AI agents — not people — evaluate your product. If you are a validator, AI helps decide which experiments reach you and whether your answers engage with the material. None of these systems is a human being, none of them is infallible, and every one of them is described below.

E.2 Where AI is used, and what it does

E.2.1 On the ideator side

• Hypothesis assistant — What it does: Asks follow-up questions about your idea and drafts a structured hypothesis (customer, problem, solution) · What it does not do: It does not evaluate whether your idea is good · Human involvement: You edit and approve every word before it is used
• Hypothesis analysis — What it does: Scores the clarity of the customer, problem and solution statements and flags risks · What it does not do: The score is a measure of how testable your statement is, not of how promising your idea is · Human involvement: Advisory only; you may ignore it entirely
• Experiment drafting — What it does: Proposes an initial set of question blocks and acceptance criteria from your hypothesis · What it does not do: It does not decide what you test · Human involvement: Every block and every criterion is editable and must be confirmed by you before launch
• Suggestions panel — What it does: Flags gaps in coverage, unbalanced questions and unrealistic framing while you build · What it does not do: It does not block you · Human involvement: Advisory only
• Free-text analysis — What it does: Reads open answers and derives structured signals, including sentiment · What it does not do: It does not read intent, honesty or emotion in any clinical sense · Human involvement: The metrics it produces feed criteria that you set
• Verdict computation — What it does: Applies the acceptance criteria you configured to the responses collected, and produces a grade and a verdict · What it does not do: It does not make a judgement about your business · Human involvement: Fully automated. Part E.5 explains how to contest a result
• Report generation — What it does: Writes the narrative report: findings, patterns, recommendations · What it does not do: It does not verify facts, cite sources or perform research · Human involvement: Fully automated. If it fails, the underlying structured data is shown instead
• Agentic panel — What it does: Instantiates AI agents that reason about your problem and then evaluate your solution against real competitors, producing a grade and a reasoning trace · What it does not do: It does not simulate human beings, does not represent a synthetic audience, and does not predict human behaviour · Human involvement: Fully automated

E.2.2 On the validator side

• Matching — What it does: Determines which validators receive an offer for a given experiment, and in what order · Effect on you: How many offers you receive · Human involvement: Automated, monitored by staff who can override it
• Profile quality check — What it does: Assesses the substance and internal consistency of the professional materials you submit at onboarding · Effect on you: Whether your application is approved, deferred or refused · Human involvement: Automated assessment, human decision on request, and always human on appeal
• Response quality check — What it does: Assesses whether a submitted response engages with the experiment · Effect on you: Whether a response is flagged for review · Human involvement: Automated flag, human decision before any rejection
• Reliability score — What it does: Aggregates your history into a single figure · Effect on you: Eligibility for experiments restricted to higher-rated validators · Human involvement: Automated, adjustable by staff, disclosed to you on request
• Fraud and abuse detection — What it does: Flags patterns indicating multiple accounts, automated answers or manipulation · Effect on you: Whether your account is reviewed or restricted · Human involvement: Automated flag, human decision

E.2.3 What never feeds these systems

Declining an offer, letting an offer expire, abandoning an experiment, giving negative or critical answers, reporting content, complaining, or exercising any right under the Terms is never an input to any of the systems above and never counts against you.

E.2.4 What we never process with AI

We do not use artificial intelligence to infer your emotional or psychological state, to analyse your private communications, to predict whether you would exercise collective rights or join a trade union, or to infer racial or ethnic origin, migration status, political opinions, religious or philosophical beliefs, disability, health, or sex life. We use biometric data only for the one-to-one identity check performed by our payment provider, with your explicit consent, and for no other purpose.

E.3 How artificially generated content is marked

Every output generated wholly or predominantly by an AI system is:

(a) visibly labelled in the interface at the point where it is shown to you; and

(b) marked in a machine-readable form, embedded in or accompanying the exported file, so that downstream systems can detect that it was artificially generated.

You must not remove, alter, obscure or degrade these markings. Doing so breaches the Terms and may breach Article 50 of Regulation (EU) 2024/1689.

If you republish AI Output to the public, the transparency obligations attaching to that publication become yours, unless you have reviewed the output and assume editorial responsibility for it.

E.4 What these systems get wrong

We would rather tell you this plainly than bury it.

E.4.1 Hallucination. Generative models produce text that is fluent and plausible whether or not it is true. A report may state a fact that is not a fact, attribute a view to a source that never held it, or reason from a premise it invented.

E.4.2 Non-determinism. The same experiment run twice may not produce identical output. Reasoning traces in particular are re-generated, not retrieved.

E.4.3 The verdict measures your criteria, not your market. A grade is the proportion of your own make-or-break criteria that were met by the responses actually collected. Set easy criteria and you will pass. Set impossible ones and you will fail. The verdict is not a prediction of commercial success and has no statistical relationship to market demand.

E.4.4 Small panels. The number of responses in a standard experiment is small. Results are not statistically representative of any population unless the experiment was specifically designed and sized to be, which the standard service does not do.

E.4.5 Reasoning traces are descriptions, not explanations. A trace tells you what the model wrote about its decision. It is not an audited account of what happened inside the model.

E.4.6 Agentic panels tell you about machines. An agentic result tells you how a large language model, given a task, treated your product relative to competitors. This is genuinely useful — increasingly, a machine sits between a product and its customer. But it tells you nothing, on its own, about how a person would react.

E.4.7 Training-data effects. Models reflect the data they were trained on, which over-represents some markets, languages, industries and viewpoints and under-represents others. Outputs may carry that skew.

E.4.8 Outputs are not unique. Another user with a similar idea may receive substantively similar output.

E.4.9 None of it is advice. No output of the Service is legal, financial, investment, tax, medical, psychological or regulatory advice.

E.5 Automated decisions, and your rights

E.5.1 When this applies. Some decisions in the Service are taken by automated means at the moment they are taken. Where such a decision produces legal effects concerning you or similarly significantly affects you, the rights in this Section apply.

E.5.2 Your rights. You have the right to:

(a) obtain human intervention — a qualified person reviews the decision;

(b) express your point of view — submit any explanation or evidence you consider relevant;

(c) obtain an explanation — in plain language: the procedure and the principles actually applied, the main categories of data used and their relative importance, and the main reasons for the outcome in your case;

(d) contest the decision and have it corrected if it was wrong.

E.5.3 Trade secrets are not a veto. Where genuine confidentiality prevents full disclosure to you, we provide the withheld information to the competent supervisory authority or court instead. We do not use confidentiality to refuse an explanation altogether.

E.5.4 How to exercise them. Write to gabri@validating.studio, stating the decision you are contesting. We complete the review and reply with individual reasons within 14 days.

E.5.5 What is never automated alone. No permanent suspension of an account, no rejection of a response, and no forfeiture of compensation already earned takes final effect by automated means alone. A human being reviews it, and in any event does so on your request.

E.5.6 Correction. If the review shows the decision was wrong, we reverse it immediately and restore what can be restored, including reinstating eligibility and releasing any suspended payment.

E.6 The models and providers behind the Service

• Generative language models (hypothesis, suggestions, reports, agentic panel, quality assessment) — Provider: Anthropic (Claude) · Location: United States, with EU-region processing where available · Notes: Contracted on terms under which content submitted through the interface is not used to train the provider's models. Transfers made under appropriate safeguards
• Database, storage, authentication, edge functions — Provider: Supabase · Location: European Union (Frankfurt) · Notes: Primary data store; row-level access control on every table
• Hosting and content delivery — Provider: Vercel · Location: European Union region, global edge network · Notes: Serves the web application and the marketing site
• Content management for the public site — Provider: Sanity · Location: European Union / global CDN · Notes: Marketing pages, blog and this notice
• Payments, payouts and identity verification — Provider: Stripe · Location: Ireland and United States · Notes: Card data and identity documents are held by the provider, never by us
• Product analytics and feature flags — Provider: PostHog · Location: European Union · Notes: EU-hosted instance
• Error monitoring — Provider: Sentry · Location: European Union / United States · Notes: No personal data in application logs; users are identified by opaque identifiers
• Transactional email — Provider: Resend · Location: United States · Notes: Sign-in links, receipts and service notices
• Push notifications — Provider: Firebase Cloud Messaging (Google) · Location: United States · Notes: Delivery of Offers to the validator application

The authoritative and current list, with each provider's role, is maintained in the Privacy Policy.

We call these models through their interfaces. We do not train, fine-tune or modify them. Under the AI Act we act as the provider of the AI systems we make available under our own name, and you act as their deployer in respect of your use of the output.

E.7 What we do with what you give us

E.7.1 We do not sell your content to model providers. User content is not supplied to any third-party model provider for the purpose of training that provider's models.

E.7.2 Improving our own systems. We use material generated through the Service to improve our own prompts, heuristics, evaluation criteria and quality controls, in aggregated or de-identified form wherever that is sufficient. The legal basis, and how to object, are set out in the Privacy Policy.

E.7.3 Aggregated benchmarks. We build aggregated, de-identified statistics — for example, how often experiments in a category reach a given verdict, or recurring patterns in how machine agents evaluate a category — and we may make comparative insight from them available to other users. These aggregates never identify you, your hypothesis, your solution or any individual response, and are built so that they cannot reasonably be re-identified.

E.8 Human oversight, on both sides

E.8.1 Ours. The automated systems described here are monitored by staff who have the authority, the competence and the technical ability to override their output, and who review every decision that materially affects a person when asked to.

E.8.2 Yours. You are responsible for exercising judgement over any output you use. A report is an input to a decision, not the decision. Where the decision matters — money, launch, hiring, a claim to a customer or an investor — take independent professional advice.

E.9 What we do not permit

Regardless of what a model will technically produce, the Service must not be used for any practice prohibited by Article 5 of Regulation (EU) 2024/1689, nor to generate non-consensual intimate imagery or child sexual abuse material, nor to manipulate behaviour in a way that causes harm, nor to discriminate unlawfully. The full list of prohibited content and conduct is in Annex B to the Terms.

E.10 AI literacy

We publish this Notice, the Instructions for Use of each AI system, and in-product explanations of what each system does and does not do, so that you can use them competently. We train our own staff on the same material. If something here is unclear, write to us and we will explain it.

E.11 Changes to this Notice

We update this Notice whenever we add, remove or materially change an AI system, or the way an automated decision is taken. Material changes are notified in the same way as material changes to the Terms. Previous versions are kept and are available on request from gabri@validating.studio.

E.12 Contact

• Questions about this Notice — Address: hello@validating.studio
• Human review of an automated decision — Address: gabri@validating.studio
• Data protection requests — Address: gabri@validating.studio
• Reporting content — Address: gabri@validating.studio
• Accessibility — Address: hello@validating.studio

You also have the right to lodge a complaint with a data protection supervisory authority, and, in the Netherlands, with the Autoriteit Persoonsgegevens.

This Notice is provided in an accessible format. If any part of it is inaccessible to you, tell us at hello@validating.studio and we will provide it in another form.