Acceptable Use Policy
ACCEPTABLE USE POLICY
METADAT-AI LTD | Effective: October 2026
1. PURPOSE AND SCOPE
This Acceptable Use Policy ("AUP") sets out the conditions under which METADAT-AI LTD (Company No. 17486598) ("Metadat-AI", "we", "us") permits access to and use of the Metadat-AI Platform, API, and any related services (collectively, the "Software"). It applies to all customers, end users, resellers, API developers, and any other person who accesses or uses the Software ("you", "User").
This AUP is incorporated by reference into the Customer Terms, the End User Licence Agreement (EULA), the API Terms of Use, the Reseller Agreement and any Order Form governing your access to the Software. It is the single statement of the uses of the Software that are prohibited or restricted, and those documents refer to it rather than repeating it. Capitalised terms not defined here have the meanings given in Part 1 of the Master Definitions Schedule (www.metadat-ai.com/legal/definitions). Its restrictions apply in addition to those in each of those agreements; if they conflict, the order of precedence in the relevant agreement applies.
This AUP is a Living Document. We may update it by publishing a new version at www.metadat-ai.com/legal/policies/aup. Updates take effect in accordance with section 13 of the Standard Legal Terms (www.metadat-ai.com/legal/standard-terms): an update that materially and adversely affects you takes effect 30 days after we give notice of it by email or in the Software, and you may terminate the affected agreement before then; updates required by law, a court or a Regulator, or needed to address a security risk, take effect on publication. Continued use of the Software after an update takes effect is acceptance of it.
2. PERMITTED USES
Subject to compliance with this AUP and your applicable licence agreement, you may use the Software to:
-
process video, audio, and image content for legitimate business purposes, including metadata extraction, transcription, tagging, classification, search indexing, and content analysis;
-
integrate the Software with your own applications and workflows via the Metadat-AI API, in accordance with the API Terms of Use;
-
access outputs generated by the Software for internal business, commercial, or research purposes, subject to the IP provisions of your licence agreement and clause 7 of this AUP;
-
allow Authorised Users within your organisation to access the Software in accordance with your licence agreement.
3. ABSOLUTE PROHIBITIONS
The following uses are strictly prohibited regardless of any contractual arrangement or instruction from a third party. Breach of any prohibition in this clause 3 constitutes a material breach and grounds for immediate suspension or termination:
3.1 Child safety. You must not use the Software to generate, process, store, distribute, or facilitate access to any content that sexually exploits, abuses, or endangers minors. This prohibition is absolute and admits of no exception.
3.2 Illegal content. You must not use the Software to process, create, store, or disseminate content that is unlawful under the laws of England and Wales or any jurisdiction in which you operate, including content that incites violence or hatred, constitutes a criminal offence, or infringes third-party intellectual property rights.
3.3 Weapons, terrorism, and serious harm. You must not use the Software in connection with the design, manufacture, or deployment of weapons, explosives, or other instruments of serious physical harm, or in connection with any terrorist activity or the support or financing of terrorism.
3.4 Biometric surveillance without consent. You must not use the Software to conduct facial recognition, gait analysis, or other biometric identification of individuals without a valid lawful basis and, where required by law, explicit informed consent. You must comply with all applicable law in any jurisdiction where the data subjects are located.
3.4A Age restriction for biometric submissions. You must not enrol in any gallery, or identify or match using any biometric processing feature of the Software (including facial recognition, biometric categorisation or speaker identification), any individual whom you know or reasonably believe to be under 18 years of age. The faces of such individuals may be detected in submitted content only as non-enrolled persons and must not be identified. Where you are uncertain whether an individual is under 18, you must treat that individual as under 18. There is no exception for parental or guardian consent.
3.4B Biometric sensitive attribute prohibition. You must not use the Software, any AI Outputs, or any biometric processing feature of the Software to categorise, classify, or infer information about any individual in relation to any of the following attributes: (a) race or ethnic origin; (b) political opinions; (c) trade union membership; (d) religious or philosophical beliefs; (e) sex life; or (f) sexual orientation. This prohibition applies regardless of whether such categorisation was the primary or intended purpose of the relevant query, and includes any case where an AI Output incidentally reveals or suggests such an attribute. Where you use custom prompts or otherwise instruct the Software to perform tasks that may produce such outputs, you remain solely responsible for ensuring that the Software is not used in a manner that contravenes this clause. For the avoidance of doubt, this clause does not prohibit output of biological sex (male/female) where explicitly requested, provided that such output is not used to infer or deduce any attribute listed at (a) to (f) above, and provided that you remain responsible for compliance with all other applicable laws (including the Equality Act 2010 and equivalent anti-discrimination legislation) governing any downstream use of that output.
3.4C Age and gender estimation. This clause applies to any estimate of a person’s age or gender produced by the Age and Gender Estimation Features or any other feature of the Software. Any such estimate is an estimate only, with a margin of error. You must: (a) treat it as an estimate and not as a determination of the person’s age or gender; (b) verify a person’s age by appropriate independent means before treating that person as an adult or as a child for any purpose; and (c) never use an estimate as the sole basis for any decision about a person.
3.4D EU AI Act prohibited practices. Whether or not the EU AI Act applies to you, you must not use the Software for any practice prohibited by Article 5 of the EU AI Act, including creating or expanding a facial recognition database by untargeted scraping of facial images, real-time remote biometric identification in publicly accessible spaces, inferring emotions in a workplace or education institution, social scoring, and generating or manipulating realistic intimate images of an identifiable person without their consent.
3.4E Law enforcement and immigration. You must not use the Software for or on behalf of any police, law enforcement, border control, immigration or asylum authority, or for law enforcement or immigration purposes, unless Metadat-AI has expressly agreed in a signed agreement after confirming that the use meets every applicable legal requirement (including the EU AI Act, data protection law and any required authorisation).
3.5 Deceptive synthetic media. You must not use the Software to generate or manipulate audio-visual content that impersonates a real individual without their consent, or that is designed to deceive recipients as to its origin or authenticity (commonly referred to as "deepfakes").
3.6 Automated decisions. You must not use the Software or any AI Output as the sole or primary basis for: (a) any decision that produces a legal effect on, or similarly significantly affects, any individual, including decisions about employment (such as hiring, performance evaluation or dismissal), credit, insurance, housing, social benefits, education, healthcare, law enforcement, immigration, border control or judicial proceedings; or (b) real-time autonomous decisions in safety-critical contexts, such as medical diagnosis, vehicle control or critical national infrastructure. Where AI Outputs inform any such decision, a human with relevant authority and competence must review the AI Output and the individual’s circumstances, and take responsibility for the decision, before it is made.
3.7 Sanctioned persons and prohibited jurisdictions. You must not use the Software in connection with any Sanctioned Person or in any Prohibited Jurisdiction, as defined in the Metadat-AI Sanctions and Export Control Policy (www.metadat-ai.com/legal/policies/sanctions). You represent and warrant that neither you nor any of your Authorised Users is a Sanctioned Person.
3.8 Disinformation and political manipulation. You must not use the Software, any AI Output, or any feature of the Platform to generate, distribute or facilitate the creation of: (a) disinformation, synthetic news, or fabricated media that is false or misleading and likely to deceive a reasonable recipient; (b) political campaign material, electoral advertising, or content designed to manipulate voting behaviour or electoral outcomes; or (c) content targeting specific individuals or demographic groups for the purpose of political manipulation, coordinated inauthentic behaviour, or suppression of legitimate political participation.
3.9 Third-party model terms. You must not use the Software in breach of any use policy of a third-party model provider identified in the Open Source Licensing Statement (for example, Google’s prohibited use policy for its open models).
4. RESTRICTED USES
The following uses are prohibited unless Metadat-AI has given its prior written consent, and may then be subject to additional terms:
4.1 Training competing models. You must not use outputs of the Software, or any data derived from or generated by the Software, to train, fine-tune, or improve any machine learning or artificial intelligence model that competes, or is intended to compete, with the Software or any Metadat-AI product or service, without Metadat-AI's prior written consent.
4.2 Reverse engineering. You must not attempt to decompile, disassemble, reverse engineer, or otherwise derive source code or underlying model weights from the Software, or attempt to identify, extract, replicate or reconstruct any model architecture, model weights, training data or algorithm of the Software (including by systematic or high-volume querying, inference attacks, prompt injection or adversarial inputs), except to the extent expressly permitted by applicable law.
4.3 Automated decisions. See clause 3.6.
4.4 Sub-licensing and resale. You must not sub-license, resell, or make the Software available to third parties except as expressly authorised under a Reseller Agreement with Metadat-AI.
4.5 Processing of special category data. You must not submit to the Software any special category personal data (as defined in UK GDPR Article 9) or criminal offence data, except where you have (a) identified a valid Article 9 condition, (b) carried out an appropriate Data Protection Impact Assessment, and (c) agreed the processing in writing with Metadat-AI under the Data Processing Agreement. This clause does not apply to special category data that appears incidentally in media content and that you do not use the Software to process for that purpose, or to Biometric Data processed under the Biometric Addendum.
4.5A Credit, employment, insurance and sensitive identifiers. Without Metadat-AI’s prior written consent, you must not use the Software: (a) as a factor in establishing any individual’s eligibility for credit, employment or insurance, or for any purpose that would make the Software or Metadat-AI subject to the US Fair Credit Reporting Act; or (b) to collect, store or process government-issued identification numbers, health information (including Protected Health Information under the US Health Insurance Portability and Accountability Act) or financial account data (including card numbers, security codes and Non-Public Information under the US Gramm-Leach-Bliley Act).
5. TECHNICAL OBLIGATIONS
5.1 API key security. You must keep API keys and other credentials secure, must not share or publish them, and must comply with the API Terms of Use.
5.2 Rate limits and fair use. You must not circumvent, bypass or interfere with rate limits, throttling controls or other technical restrictions implemented by Metadat-AI.
5.3 No interference. You must not use the Software to introduce malware, viruses, worms, trojan horses, or other malicious code; to conduct denial-of-service or distributed denial-of-service attacks; or to gain unauthorised access to any system, network, or data.
5.4 Vulnerability disclosure. If you discover a security vulnerability in the Software, you must notify security@metadat-ai.com promptly and in confidence, and must not disclose the vulnerability publicly until Metadat-AI has had a reasonable opportunity to investigate and remediate it.
5.5 Content submission standards. You must not submit to the Software any content that is encrypted in a manner that prevents Metadat-AI from processing it as required, or that is designed to cause the Software to malfunction or produce erroneous outputs.
5.6 Custom Prompts. Where the Software permits you to submit custom instructions, queries, or prompts to configure, direct, or extend the Software’s AI processing (“Custom Prompts”), you are solely responsible for the content and lawfulness of those Custom Prompts in every jurisdiction in which you operate. You must ensure that Custom Prompts comply with this AUP and all applicable law.
6. DATA PROTECTION AND PRIVACY
6.1 Data Processing Agreement. Processing of personal data through the Software is governed by the Data Processing Agreement (www.metadat-ai.com/legal/dpa). You must not submit personal data until it applies to you.
6.2 Children's data. Where content you submit may include personal data of individuals under 18, you must have a lawful basis and any authorisation required by law for submitting it. This does not permit any processing prohibited by clause 3.4A.
6.3 Data minimisation. You must not submit personal data to the Software beyond what is necessary for the legitimate processing purpose. Where the same processing outcome can be achieved using anonymised or pseudonymised data, you should do so.
7. AI OUTPUTS
7.1 Human review. AI Outputs are probabilistic and never fully accurate. You must not rely on an AI Output as the sole basis for any decision, and a human must review AI Outputs before any decision is made using them (see also clause 3.6).
7.2 Disclosure and representation. Where applicable law or professional regulation requires you to disclose that content or a decision was generated or assisted by artificial intelligence, you must make that disclosure. You must not represent any AI Output as produced by a human, or as approved, vetted, endorsed or certified by Metadat-AI, where that would be false or misleading.
7.3 Responsible AI. Your use of the Software is subject to Metadat-AI's Responsible AI Policy (www.metadat-ai.com/legal/policies/responsible-ai). You agree to use the Software in a manner consistent with that policy, including avoiding uses that are discriminatory, deceptive, or harmful to individuals or society.
8. REPORTING VIOLATIONS
If you become aware of any use of the Software in violation of this AUP, you should report it to us at legal@metadat-ai.com. Reports should include sufficient detail to allow us to investigate, and will be treated as confidential. We aim to acknowledge receipt within five (5) Business Days and will investigate as we consider appropriate.
If you are a reseller and become aware of a violation by one of your customers, you must notify us as soon as practicable and cooperate with our investigation. Your obligations under this clause are in addition to, not in substitution for, your obligations under the Reseller Agreement.
In addition to the reporting obligation above, you must notify Metadat-AI promptly in writing at legal@metadat-ai.com if you receive any third-party claim, demand, notice or complaint alleging that your use of the Software or any AI Output infringes a third party's intellectual property rights, privacy rights, image rights, or any other legal right. Notifications under this clause must include sufficient detail of the claim and the relevant use of the Software to allow Metadat-AI to assess the position.
8A. CONDUCT IN INTERACTIONS WITH METADAT-AI
You must treat Metadat-AI’s personnel with courtesy and respect in all communications, including by telephone, email, messaging, video conferencing, and any support channel. Metadat-AI will not tolerate bullying, harassment, threats or abusive or intimidatory conduct directed at its personnel. Metadat-AI reserves the right, without prejudice to any other right or remedy, to suspend access to the Software with immediate effect where such conduct occurs, and to treat such conduct as a material breach of this AUP entitling Metadat-AI to terminate the applicable agreement on written notice.
9. CONSEQUENCES OF BREACH
9.1 Suspension. Metadat-AI may immediately suspend your access to the Software, without prior notice, if it reasonably believes you have committed or are committing a breach of clause 3 (Absolute Prohibitions) or are causing, or are likely to cause, material harm to Metadat-AI, other users, or third parties. Metadat-AI aims to notify you as soon as practicable after suspension and to give you an opportunity to respond.
9.2 Termination. Metadat-AI may terminate the applicable agreement (such as the Customer Terms, the EULA or the API Terms of Use) with immediate effect upon written notice if you commit a material breach of this AUP that is incapable of remedy, or that you fail to remedy within ten (10) Business Days of written notice requiring remedy.
9.3 Reporting to authorities. Metadat-AI may report breaches to relevant regulatory authorities where required or appropriate.
10. CONTACT