Trust & Safety

Formless Inc., doing business as Vidzap, operates a platform that generates synthetic video and audio of people. That capability can be misused, and pretending otherwise would not make anyone safer. This policy sets out what we do about it.

It works alongside our Acceptable Use Policy, which defines what is prohibited, and our Media Upload Agreement, which governs uploads of real people.

1. Our approach

We combine three layers:

• Automated detection at upload and at generation — hash matching against known illegal imagery, classifiers for prohibited categories, and prompt-level filtering.
• Human review for flagged content, escalations and anything an automated system is not confident about.
• User reporting, because no detection system catches everything, and the person depicted is often the first to know.

We would rather be proportionate than maximal. Most policy issues are handled with a block or a warning. The categories below are the ones where we are not proportionate at all.

2. Child safety

We have zero tolerance for child sexual abuse material (CSAM) and for the sexualization of minors, real or synthetic.

• No media depicting a minor may be uploaded to any Vidzap feature, for any purpose.
• No prompt may be used to generate a person who appears to be a minor in a sexual, suggestive or intimate context.
• We deploy hash-matching and classifier-based detection at upload and at generation.
• Confirmed CSAM is removed immediately, the account is terminated without notice, and the content and associated account records are preserved.
• We report apparent CSAM to the National Center for Missing & Exploited Children (NCMEC) as required by 18 U.S.C. § 2258A, and we cooperate with resulting law-enforcement process.

There is no appeal on a confirmed CSAM determination.

To report child safety content, email trust@vidzap.ai with “URGENT — CHILD SAFETY” in the subject. If a child is in immediate danger, contact local law enforcement first. In the US you can also report to the NCMEC CyberTipline.

3. Non-consensual intimate imagery

Creating or distributing intimate, nude or sexually explicit depictions of a real, identifiable person without their documented consent is prohibited. This covers synthetic imagery, “nudified” or undressed versions of clothed photographs, and content framed as fiction but recognizably depicting a real person.

• Verified reports are actioned within 48 hours, and usually much faster.
• You do not need to prove your identity to report imagery of yourself — we will work with what you can safely provide, and we will not require you to send us the material itself.
• Removed content is hashed and shared into industry programs so it cannot easily be re-uploaded here or elsewhere.
• Accounts responsible are terminated.

Report to trust@vidzap.ai. If you are in distress, StopNCII.org and the Cyber Civil Rights Initiative offer support and can help with cross-platform removal.

4. Impersonation and harmful deepfakes

You may not generate content depicting a real, identifiable person without the consent required by the Media Upload Agreement. Public figures are not an exception.

We act on:

• synthetic content depicting a real person saying or doing something they did not;
• content falsely presenting itself as from, endorsed by or affiliated with a real person, brand, government body or news organization;
• voice clones of real people made without consent;
• fabricated footage presented as authentic news, documentary, security or bodycam material; and
• synthetic depictions of candidates, officials or electoral processes intended to deceive.

If someone has generated content depicting you, report it to trust@vidzap.ai. You do not need to be a Vidzap user to report, and you do not need a lawyer.

5. Provenance and AI disclosure

Every Output is synthetic, and we think it should be identifiable as such.

• We attach machine-readable provenance signals — content credentials and, where applicable, watermarking — to Outputs where technically feasible.
• You may not remove, obscure, strip or falsify those markings. Doing so is a breach of our Terms of Service.
• Where the law or an advertising platform requires you to disclose that content is AI-generated — including disclosure of synthetic endorsers and testimonials — that obligation is yours as the publisher. Under the EU AI Act, deployer-side deepfake disclosure sits with you.

Provenance technology is imperfect. Signals can be lost through re-encoding, cropping or platform processing, and their absence is not proof that content is authentic.

6. How to report

Please include what the content is, where you saw it, why it breaches our rules, and how we can reach you. Do not attach illegal material to a report — describe it and give us a location instead.

Reports made in bad faith, or used to harass another user, are themselves a policy breach.

7. What we do about violations

Depending on severity, intent and history:

1. Block — the generation is refused before it runs.
2. Remove — the content is deleted and cannot be re-rendered.
3. Warn — a formal notice, with the policy cited.
4. Restrict — specific features are disabled, or the account is rate-limited.
5. Suspend — access is paused pending review.
6. Terminate — the account is closed and unused credits forfeited.
7. Report — records preserved and referred to law enforcement.

Immediate termination, without warning, applies to: confirmed CSAM; non-consensual intimate imagery; and sanctions or trade-control violations.

8. Appeals

If you think we got it wrong, email trust@vidzap.ai within 30 days of the action, with your account email, what was actioned, and why you believe it was a mistake.

• Appeals are reviewed by a person who was not involved in the original decision.
• We aim to respond within 14 days.
• If we agree we were wrong, we restore the content or the account and remove any associated strike.
• CSAM determinations are not appealable.

Users in the EU have additional out-of-court dispute settlement options under the Digital Services Act, and we will identify the relevant body in our statement of reasons.

9. Law-enforcement requests

Law enforcement should direct legal process — subpoenas, court orders, warrants and preservation requests — to legal@vidzap.ai, on official letterhead, with a return address at the requesting agency’s domain.

We review each request against applicable law, disclose only what the process requires, and notify the affected user unless we are legally prohibited from doing so or notification would create a risk of harm. In an emergency involving a risk of death or serious physical injury, mark the request EMERGENCY DISCLOSURE REQUEST and describe the nature of the emergency.

We publish nothing about individual accounts outside of valid legal process.

10. Cooperation and standards

We work with NCMEC on child-safety reporting, participate in industry hash-sharing programs for known CSAM and non-consensual intimate imagery, and align our practices with US federal reporting obligations, the EU AI Act’s transparency requirements, and the EU Digital Services Act’s notice, statement-of-reasons and appeal mechanisms.

Our systems are not perfect, and neither are our judgement calls. If you think a rule here is wrong, or that we are enforcing it badly, tell us at trust@vidzap.ai.

11. Contact

Formless Inc. (DBA Vidzap)
Trust & Safety: trust@vidzap.ai
Legal and law enforcement: legal@vidzap.ai
Copyright and trademark: dmca@vidzap.ai
Mailing address: 18117 Biscayne Blvd, Suite #2339, Miami, FL 33160