coub.ai

Acceptable Use Policy

Effective date: June 16, 2026 Operating entity: [Coub.AI Ltd, Seychelles — confirm legal entity & registered address] Service: coub.ai — a multi-modal AI generation platform (image, video, audio, and 3D), currently in closed beta.

This Acceptable Use Policy ("AUP" or "Policy") governs your use of coub.ai (the "Service") operated by [Coub.AI Ltd, Seychelles — confirm legal entity & registered address] ("Coub.AI," "we," "us," or "our"). It is incorporated by reference into our Terms of Service and forms part of your agreement with us. Capitalized terms not defined here have the meaning given in the Terms of Service.

This Policy applies to everyone who accesses or uses the Service, including beta participants, and to all content you submit (your "Inputs," including prompts, reference images, audio, and other uploads) and all content the Service generates for you ("Outputs"). For moderation purposes, "Content" means Inputs and Outputs together.

By using the Service you agree to this Policy. If you do not agree, do not use the Service.

1. Summary (non-binding)

This section is a plain-language overview. The numbered clauses below are what actually binds you; where they conflict with this summary, the clauses control.

  • Do not create, upload, or request child sexual abuse material (CSAM). There is zero tolerance, and we report it to authorities.
  • Do not make sexual content involving minors or minor-coded characters, non-consensual intimate imagery, or deepfakes of real, identifiable people without their consent.
  • Do not use the Service for malware, phishing, fraud, weapon or explosive construction instructions, hate speech, or harassment.
  • Do not abuse free credits, run multiple accounts to farm credits, scrape the Service, evade rate limits, strip watermarks, or resell our Outputs as a competing generation service.
  • We use automated moderation and human review. Violations can lead to flagging, suspension, or a permanent ban. CSAM is referred to law enforcement.

2. Scope and relationship to other terms

2.1 This Policy supplements, and does not replace, our Terms of Service and Privacy Policy. In the event of a direct conflict on a use-and-conduct question, this Policy controls; on a commercial or liability question, the Terms of Service control.

2.2 This Policy also serves as the operational source of truth for our content-moderation rules. The categories, thresholds, and enforcement steps described here are the rules our automated systems and reviewers apply.

2.3 We may generate Outputs using third-party AI providers (including fal.ai and the underlying model vendors). Your use of Outputs is additionally subject to those providers' acceptable-use and licensing terms. Where their terms are stricter than ours, the stricter rule applies to the affected model or Output.

3. Eligibility, accounts, and regional restrictions

3.1 Minimum age. You must be at least 18 years old to use the Service. The Service is not directed to children, and we do not knowingly permit access by anyone under 18. If we learn that an account belongs to a minor, we will suspend it.

3.2 One person, accurate information. You must register with accurate information and may operate the account only as the individual or entity that registered it. Sharing, selling, or transferring accounts is not permitted.

3.3 Closed beta. The Service is currently a closed, invite-based beta. Access may be granted, limited, or revoked at our discretion. Features, credit balances, and limits may change without notice during the beta.

3.4 Sanctioned regions. The Service is not available to users located in, ordinarily resident in, or accessing from jurisdictions subject to comprehensive sanctions, currently Iran, North Korea (DPRK), Cuba, and Syria. We block these regions at signup and at content delivery. You may not use a VPN, proxy, or other means to circumvent this restriction. You also may not use the Service if you are on an applicable denied-party or sanctions list.

3.5 We may restrict or refuse access from any region, network, or account as required by law or to protect the integrity of the Service.

4. Prohibited content

You may not create, request, upload, generate, store, display, distribute, or attempt to generate any of the following through the Service. This applies to Inputs (including prompts and reference uploads) and to Outputs.

4.1 Child sexual abuse material (CSAM) — zero tolerance

4.1.1 You may not create, request, upload, solicit, or attempt to generate any content that depicts, describes, or sexualizes a minor, including:

  • real, AI-generated, photorealistic, illustrated, cartoon, anime, or "minor-coded" depictions of persons who are or appear to be under 18;
  • sexualized depictions of characters that are childlike in appearance, body, or context, regardless of any claimed or stated age;
  • "ageplay," "loli/shota," or comparable themes; and
  • uploading images of real children as reference material for sexual or sexualized Outputs.

4.1.2 Zero tolerance and reporting. CSAM detection results in immediate, permanent account termination without warning and without refund. We preserve the relevant Content and account records and report confirmed or suspected CSAM to the U.S. National Center for Missing & Exploited Children (NCMEC) and/or other competent authorities as required or permitted by law. Cloudflare CSAM scanning runs on stored objects (R2); matches are escalated under this clause. We cooperate with lawful requests from law enforcement.

4.1.3 No "artistic," "fictional," "satirical," or "research" justification is accepted for content under this section.

4.2 Non-consensual and intimate-image abuse

4.2.1 You may not create or distribute intimate, nude, or sexually explicit imagery of an identifiable real person without that person's clear consent ("non-consensual intimate imagery," NCII), including:

  • "nudify"/undress transformations of clothed photos of real people;
  • placing a real, identifiable person's face or likeness into sexual or intimate Outputs; and
  • revenge-porn, blackmail, or sextortion material.

4.2.2 Sexual content depicting a real, identifiable person always requires that person's consent. You are responsible for holding that consent and must provide evidence of it on request.

4.3 Impersonation, deepfakes, and likeness

4.3.1 You may not create Outputs that realistically depict a real, identifiable person (whether public or private) in a way that a reasonable viewer could mistake for genuine, where you do not have that person's consent and the depiction is deceptive, defamatory, sexual, or harmful.

4.3.2 Private individuals. You may not generate deepfakes or realistic likenesses of identifiable private individuals without their consent.

4.3.3 Public figures and celebrities. You may reference public figures only for lawful commentary, parody, or satire that is clearly not presented as real, and never in a way that is sexual, defamatory, fraudulent, or that falsely attributes statements, endorsements, or actions to them. You may not use a celebrity or public figure's likeness or voice to imply endorsement, to deceive, or for commercial exploitation of their identity.

4.3.4 You may not impersonate Coub.AI, our staff, another user, or any organization, including by generating fake official communications, credentials, or identity documents.

4.4 Sexual content generally

4.4.1 Any sexual or adult content permitted on the Service (where and to the extent it is permitted at all) is subject to all rules in this Section 4, in particular 4.1, 4.2, and 4.3. Where applicable AI-provider terms prohibit adult or sexual content for a given model, that prohibition applies and overrides any narrower reading of this Policy.

4.4.2 You may not generate sexual content involving non-consenting persons, coercion, or anyone who is or appears to be a minor (see 4.1).

4.5 Violence, gore, and harm

4.5.1 You may not generate content that depicts extreme, gratuitous, or realistic violence or gore intended to shock, glorify, or incite harm; that promotes, instructs, or encourages self-harm or suicide; or that depicts real, identifiable victims of violence or tragedy in a degrading way.

4.6 Hate, harassment, and abuse

4.6.1 You may not generate content that attacks, demeans, dehumanizes, or incites hatred or violence against people based on a protected characteristic (including race, ethnicity, national origin, religion, caste, sex, gender, gender identity, sexual orientation, disability, age, or serious disease), or that promotes hateful ideologies or symbols.

4.6.2 You may not use the Service to harass, bully, threaten, dox, or target a specific person or group, including by generating images of an identifiable person to demean or intimidate them.

4.7 Malware, phishing, and security abuse

4.7.1 You may not generate content that facilitates malicious software, phishing, or credential theft, including imagery of fake login pages, spoofed brand or bank interfaces, fake security warnings, counterfeit official documents, QR codes or interfaces designed to deceive, or visual assets intended for use in phishing or social-engineering campaigns.

4.8 Weapons, explosives, and dangerous goods

4.8.1 You may not use the Service to obtain or distribute instructions, schematics, diagrams, or step-by-step guidance for manufacturing, assembling, or deploying weapons, firearms, 3D-printed weapons, explosives, or chemical, biological, radiological, or nuclear (CBRN) materials. Fictional or stylized depictions of weapons are not by themselves prohibited; operational instructions are.

4.9 Fraud, scams, and deception

4.9.1 You may not use the Service to create content for fraud, scams, or deceptive schemes, including counterfeit currency or documents, forged identification, fake certificates or diplomas, deceptive advertising, pyramid or "get-rich-quick" schemes, or any content designed to mislead people into parting with money, credentials, or personal data.

4.10 Intellectual-property infringement

4.10.1 You may not use the Service to infringe the copyright, trademark, patent, trade-secret, publicity, or other rights of any party, including by uploading reference material you do not have rights to, reproducing protected characters, brands, or logos in a way that infringes or that falsely implies affiliation or endorsement, or deliberately copying an identifiable living artist's name or distinctive style to pass Outputs off as theirs.

4.10.2 You are responsible for ensuring you have all necessary rights in your Inputs and for your use of Outputs. See our DMCA / IP complaint process in Section 9.

4.11 Other illegal or harmful use

4.11.1 You may not use the Service for any activity that is illegal in your jurisdiction or in the Seychelles, including the sale or facilitation of illegal drugs or controlled goods, human trafficking, terrorism or violent extremism, or any conduct that violates applicable sanctions or export-control law.

5. Prohibited conduct

Beyond content, the following uses of the Service are prohibited.

5.1 Abuse of free credits and multi-accounting. You may not create or operate multiple accounts to obtain additional free or promotional credits, evade limits, or evade enforcement. Creating accounts using disposable, automated, or false identities to farm free credits is prohibited. We may consolidate, void credits on, or terminate accounts we determine are linked for this purpose.

5.2 Automated access and scraping. You may not scrape, crawl, harvest, or systematically extract content, model lists, metadata, or other data from the Service except through interfaces we expressly provide for that purpose. You may not access the Service through unauthorized automated means, headless browsers, or bots.

5.3 Rate-limit and quota circumvention. You may not evade, disable, or circumvent rate limits, credit accounting, quotas, queueing, or other usage controls, including by parallelizing requests across accounts, networks, or IP addresses to exceed your allotted usage.

5.4 Watermark integrity. Free-tier Outputs carry a mandatory watermark. You may not remove, crop out, obscure, alter, or otherwise defeat any watermark, label, or provenance marker we apply, and you may not represent watermarked Free-tier Output as unwatermarked or as produced on a paid tier.

5.5 Reselling and competing services. You may not resell, sublicense, redistribute at scale, or expose the Service or its Outputs through your own product as a substitute generation, model-access, or inference service, or otherwise use the Service to build or train a competing AI generation service. Normal use of your own Outputs in your own projects is permitted within your tier's terms; operating a pass-through "generation API" or reselling raw generation capacity is not.

5.6 Security probing and interference. You may not probe, scan, or test the vulnerability of the Service or its infrastructure without our prior written authorization; breach or circumvent authentication, access controls, or sandboxing; interfere with or disrupt the Service, its servers, or networks (including via denial-of-service, flooding, or resource-exhaustion); access another user's account, Inputs, or Outputs; or attempt to extract model weights or reverse-engineer non-public components. Good-faith security research is welcome only under a separate, written arrangement with us (contact legal@coub.ai).

5.7 Prompt-based evasion. You may not use obfuscation, encoding, "jailbreak," role-play framings, indirect descriptions, or piecemeal generation to produce content prohibited by Section 4 or to defeat our moderation. Attempts to do so are themselves violations.

5.8 Misrepresentation of Outputs. Where required by applicable law or platform rules, you are responsible for disclosing that Outputs are AI-generated. You may not present Outputs as authentic photographs, recordings, or documents in a way that is deceptive or unlawful.

6. Inputs, prompts, and personal data

6.1 You are responsible for your Inputs. You represent that you have the rights and any required consents to submit your Inputs and to generate the Outputs you request, and that doing so does not violate this Policy or any law.

6.2 Prompts are sent to providers. To generate Outputs, your prompts and reference uploads are transmitted to our AI providers (including fal.ai and the relevant model vendors, and, for prompt enhancement, Anthropic) substantially as you submit them. We cannot control, redact, or recall the content of a prompt once it is sent to a generation provider.

6.3 Do not paste sensitive personal data. Because of 6.2, you should not include in prompts any personal data, confidential information, or special-category data (such as health, biometric, financial, or government-identifier data) about yourself or others that you would not want disclosed to those providers. Submitting another person's personal data in a prompt without a lawful basis may violate this Policy and applicable data-protection law.

6.4 What we do protect. We redact detected personal data from our own application logs and analytics, and we limit internal access to Content. This redaction applies to our logs and observability pipeline only; it does not and cannot remove personal data from the prompt content delivered to the generation provider.

7. Moderation, monitoring, and how we enforce

7.1 Automated moderation. We use automated systems to screen Inputs and Outputs, currently including Google (Gemini) and OpenAI omni-moderation classifiers for policy categories and Cloudflare CSAM scanning on stored objects (R2). Content may be blocked before generation, withheld after generation, or removed after the fact.

7.2 Human review. Flagged Content and accounts may be reviewed by trained staff. We aim to apply context (for example, distinguishing news, education, or satire from prohibited use) where the category permits; the categories in 4.1 admit no such exceptions.

7.3 Trust levels and flagging. Accounts carry an internal trust signal. Policy hits lower it; repeated or severe hits trigger tighter automated thresholds, manual review, throttling, or feature restrictions before any suspension.

7.4 No guarantee. Automated moderation is imperfect and may both over-block and under-block. Blocking of an Output is not legal advice that the requested content would otherwise be lawful, and allowing an Output is not our endorsement or a representation that it is lawful for your use.

7.5 Limited monitoring. We do not pre-screen all Content for purposes beyond the moderation described here, and we are not obligated to monitor use of the Service, but we may review Content and account activity to operate, secure, and enforce the Service.

8. Consequences of violations

8.1 Depending on the nature and severity of a violation, your history, and the risk presented, we may take one or more of the following actions, with or without prior notice:

  • block or remove the offending Input or Output;
  • issue a warning and lower your account trust level;
  • apply rate limits, feature restrictions, or queue deprioritization;
  • void or reverse credits obtained or used in violation of this Policy;
  • suspend the account temporarily; or
  • permanently terminate the account and associated accounts.

8.2 Severity tiers (guidance).

  • Minor / first-time, lower-risk: warning, content removal, trust reduction.
  • Repeated or moderate: temporary suspension, credit forfeiture, restrictions.
  • Severe (e.g., NCII, deepfake abuse, malware, fraud, security attacks) or repeat after suspension: permanent ban.
  • CSAM (Section 4.1): immediate permanent termination, preservation of records, and referral to NCMEC and/or law enforcement, regardless of history.

8.3 No refunds for enforcement. Where the Service is paid, credits or balances forfeited due to a violation are non-refundable. During the closed beta, credits are granted manually and have no cash value; suspension or termination does not entitle you to compensation for unused credits.

8.4 Preservation and referral. For serious violations, particularly CSAM and credible threats of harm, we may preserve relevant Content, logs, and account data and disclose them to competent authorities as required or permitted by law.

8.5 Evasion after enforcement. Creating a new account, or using another account, to evade a suspension or ban is itself a violation and grounds for terminating all related accounts.

9. Reporting abuse and IP complaints

9.1 Reporting prohibited content or conduct. If you encounter Content or behavior that violates this Policy, report it to legal@coub.ai with the relevant details (description, account or Output reference, and date/time where known). For CSAM or imminent threats to a person's safety, contact local law enforcement first, then notify us at legal@coub.ai.

9.2 NCII / deepfake takedown. If you are depicted in non-consensual intimate imagery or in a deepfake created without your consent, contact legal@coub.ai or privacy@coub.ai. We prioritize these reports, will remove confirmed material, and may act against the responsible account.

9.3 Copyright and trademark complaints. Send IP infringement notices to dmca@coub.ai. A complete notice should identify the protected work or mark, the allegedly infringing Output or material and where it appears, your contact information, a statement of good-faith belief that the use is unauthorized, a statement under penalty of perjury that the information is accurate and that you are the rights holder or authorized to act, and your signature (physical or electronic). We may remove or disable access to material that is the subject of a valid notice and may terminate repeat infringers.

9.4 Privacy and data-subject requests. For access, export, correction, or erasure of your personal data, use /privacy/data-request or contact privacy@coub.ai. See Section 11.

10. Appeals

10.1 If your Content was blocked or your account was restricted, suspended, or terminated and you believe this was a mistake, you may appeal by emailing legal@coub.ai (or replying to the enforcement notice) within 30 days. Include your account identifier, the action you are appealing, and why you believe it was incorrect.

10.2 We aim to acknowledge appeals within a reasonable time and to have them reviewed by someone other than any automated system that made the original decision, where feasible. We may uphold, reverse, or modify the action.

10.3 No appeal for CSAM. Decisions under Section 4.1 (CSAM) are final and are not subject to appeal, given the legal obligations involved.

11. Privacy, sub-processors, and compliance posture

11.1 Where we operate. Coub.AI is operated by an entity incorporated in the Seychelles. We are not directly subject to the EU AI Act or the EU Digital Services Act. Notwithstanding that, we voluntarily honor core GDPR data-subject rights for our EU users, including data export and erasure, which are implemented in the Service with a target service level of 30 days from a verified request.

11.2 Sub-processors. We use the following categories of sub-processors to operate the Service. Inputs and/or Outputs may be processed by them for the stated purpose:

  • Supabase — database and authentication (hosted in Frankfurt, EU);
  • Cloudflare — CDN, R2 object storage, WAF, and CSAM scanning;
  • fal.ai — AI generation (and the underlying model vendors it routes to);
  • Google — content moderation and embeddings;
  • OpenAI — content moderation;
  • Anthropic — prompt enhancement;
  • Upstash — caching and rate-limiting;
  • Inngest — background job processing;
  • Sentry, Better Stack, and PostHog — error monitoring and product analytics; and
  • Loops — transactional email.

11.3 Prompts and reference uploads are transmitted to generation and prompt-enhancement providers as described in Section 6. Our handling of personal data, retention periods, and the full sub-processor list are detailed in our Privacy Policy, which governs in case of any inconsistency on data-protection specifics.

11.4 We may update the sub-processor list as our infrastructure changes; material changes are reflected in the Privacy Policy.

12. Content ownership and license

12.1 Your ownership. As between you and us, you own the Outputs you generate, to the extent ownership is available and subject to the terms and licenses of the AI provider and underlying model used to create them. Some providers grant only a license, not ownership, in their outputs; your rights in an Output are no greater than what the applicable provider grants.

12.2 License to us. You grant Coub.AI a worldwide, non-exclusive, royalty-free license to host, store, cache, reproduce, transmit, display, and serve your Inputs and Outputs as needed to operate, secure, maintain, and provide the Service to you (including delivery via our CDN and storage of generation history). This license ends when the relevant Content is deleted from the Service, except for residual copies in backups and as needed to comply with law or enforce this Policy.

12.3 Watermark. Free-tier Outputs are delivered with a watermark as described in 5.4; the license in 12.2 does not entitle you to remove it.

12.4 We do not claim ownership of your Inputs or Outputs beyond the license in 12.2.

13. Beta status and changes to the Service

13.1 The Service is provided during the closed beta on an "as is" and "as available" basis. Features, models, limits, credit grants, and this Policy's enforcement thresholds may change as the beta evolves.

13.2 We may add, remove, or change available models. Because Outputs depend on third-party models, model behavior, availability, and associated terms may change without notice.

14. Changes to this Policy

14.1 We may update this Policy from time to time. When we make changes, we will revise the "Last updated" date in the disclaimer block and the "Effective date" above, and, for material changes, we will provide reasonable notice (for example, by email to the address on your account or by an in-product notice) before they take effect.

14.2 Your continued use of the Service after a change becomes effective constitutes acceptance of the updated Policy. If you do not agree to a change, you must stop using the Service.

15. Governing law and contact

15.1 Governing law and venue. This Policy is governed by the laws of [governing-law jurisdiction & venue — confirm with counsel (e.g., Seychelles)], without regard to conflict-of-laws rules, except where mandatory consumer-protection or data-protection law of your place of residence applies.

15.2 Contact.

  • General and abuse/enforcement: legal@coub.ai
  • Privacy and data requests: privacy@coub.ai or /privacy/data-request
  • Copyright / trademark (DMCA): dmca@coub.ai
  • Product support: via the in-product support channel

15.3 Entity. [Coub.AI Ltd, Seychelles — confirm legal entity & registered address].

End of Acceptable Use Policy.