Research
There is no US federal AI certificate — how we explain FTC, NIST, and state law
Why there is no federal AI product certificate, NIST RMF self-documentation, FTC disclosures. No fake certification claims.
Public hub: US AI compliance · AI certification & governance
Why we built this
The United States has no federal AI product certificate like EU CE marking. Saying “US AI Act certified” would be false. We separate binding duties from voluntary frameworks so journalists and users are not misled.
How we do it
| Layer | Nature | XenLook |
|---|---|---|
| FTC §5 | Anti-deception enforcement | AI labels · ban fake badges |
| NIST AI RMF | Voluntary risk framework | Govern·Map·Measure·Manage evidence (not NIST Certified) |
| State laws (e.g. Texas TRAIGA) | Patchwork | Prohibited-use check + RMF defense posture |
| ISO/IEC 42001 | Optional third-party cert | Gap draft only — no Certified until issued |
Public summary: /compliance/us.
Service scope
XenLook is a global service. Beta opens Korea-first. We support 12 UI locales. Foreign and U.S. residents cannot be fully blocked, so we keep a Texas TRAIGA defense posture (NIST RMF + prohibited-use checks). C2PA CA is Korea then Japan — not a global CA service. Compliance is run in-house (evidence packs, disclosures, and audits). Scope: global product · Korea beta · C2PA CA Korea→Japan.
What gets better
- Clear, honest language instead of fake certificates
- Reuse of existing Companion safety, C2PA, and disclosures
- ISO 42001 only if leadership green-lights an audit
Forbidden claims: NIST Certified · US AI Act certified · federal AI approval.