AI Likeness & Consent Policy

memacta's Persona feature lets you train an AI model on photographs of a real person to produce consistent, personalised Generations. This capability is powerful — and therefore carries significant legal and ethical obligations. This policy describes what is permitted, what is prohibited, and how we respond to complaints.

1. Permitted Uses

• Train a Persona on photographs of yourself and generate content featuring your own likeness.
• Train a Persona on photographs of another adult person for whom you hold documented, revocable, written consent that covers AI-generated content — and only use that Persona in ways the consent covers.
• Generate stylised, non-photorealistic art, avatars, or character designs inspired by (but not identifying) a real person, provided the output is clearly artistic and not likely to be mistaken for a real photograph or video of that person.

2. Prohibited Uses

You may not use memacta to:

• Generate content featuring celebrities, public figures, or politicians without their documented consent — regardless of how realistic or stylised the output is. Likeness rights (right of publicity) are recognised under California Civil Code § 3344, Texas Civil Practice & Remedies Code § 143, the Tennessee Personal Rights Protection Act, and EU General Data Protection Regulation Art. 9.
• Generate content featuring anyone other than yourself unless you possess and can produce written consent from that person covering AI-generated imagery.
• Generate content depicting minors (under 18) in any context. Persona photos flagged by our age-estimation pipeline as depicting a minor are automatically rejected.
• Generate sexually explicit content featuring any identifiable person — including yourself — unless you are on a plan tier where explicit content is explicitly enabled (currently not available on any tier) and you have provided a verified age attestation.
• Use Generations for fraud, impersonation, defamation, or harassment — including creating fake statements, fabricating events, or placing a real person in a false context intended to deceive.
• Create non-consensual intimate imagery (NCII / "deepfake porn") of any real person. This violates our Terms, is illegal in an increasing number of jurisdictions, and will result in immediate account termination and reporting to relevant authorities.

3. Regulatory Compliance

EU AI Act — Deepfake Disclosure

All content generated on memacta is marked with a C2PA (Coalition for Content Provenance and Authenticity) provenance signal embedded in the file metadata, indicating it was produced by an AI system. Free-tier downloads additionally carry a visible watermark. This satisfies the EU AI Act (Regulation 2024/1689) Article 50 requirement that providers of AI systems capable of generating synthetic media disclose that content is artificially generated.

US State AI Likeness Statutes

memacta is designed to comply with AI likeness and digital replica laws including:

• Texas: Relating to the use of an individual's likeness in synthetic media (HB 4337 / SB 2037 — 2023 session and amendments).
• California: AB 602, AB 1228, SB 1047 and related legislation governing AI-generated depictions and performer rights.
• Tennessee: ELVIS Act (Ensuring Likeness Voice and Image Security, 2024) — protecting recording artists and performers from AI vocal/likeness replication without consent.
• Federal (US): We monitor proposed federal legislation including the NO FAKES Act and the DEFIANCE Act and will update our policies as these become law.

This list is not exhaustive. Additional state and national laws apply. If your jurisdiction has enacted AI likeness legislation not listed here, contact legal@memacta.ai.

4. Reporting Unauthorised Use of Your Likeness

If you believe someone has generated content using your likeness on memacta without your consent, you have two paths:

5. How We Enforce This Policy

• At upload: Reference photos are screened for estimated age, NSFW content, and known celebrity face signatures. Photos that fail these checks are rejected before a Persona is created.
• At generation: Prompts are moderated for names of public figures, harassment patterns, and explicit content markers.
• On report: Verified likeness complaints result in immediate content removal, review of the offending account, and (for severe violations) referral to law enforcement.
• Repeat violators: Accounts with two or more substantiated likeness violations are permanently terminated. We cooperate fully with law-enforcement investigations related to non-consensual intimate imagery.