Artificial Intelligence Accountability Act
New Mexico House Bill 141 (2026) was introduced but not enacted; the official record shows it was "Action Postponed Indefinitely" on 2026-01-22. If enacted, it would create the Artificial Intelligence Accountability Act (Sections 1-9). Covered providers are generative-AI services with more than 2 million monthly U.S. users. Under Section 3, they would have to offer users an option for a manifest AI disclosure on synthetic content. They would also have to embed a hard-to-remove latent disclosure that names the provider, the system and version, the date and time, and a unique identifier. Providers that license their systems to third parties would have to require compliance by contract and revoke the license within 96 hours of discovering material noncompliance. Under Section 4, covered providers would have to offer a free public provenance detection tool that reveals system provenance data only, not personal provenance data. Section 5 would require capture device manufacturers, for devices first produced for sale in the state on or after January 1, 2028, to embed a latent disclosure by default and let users opt out. Section 6 would require large online platforms (more than 2 million monthly U.S. users) to detect and display system provenance data and not strip compliant provenance data or digital signatures. They would also have to offer a takedown interface for deceptive synthetic content and respond within 72 hours. Section 6 also gives platforms a limited good-faith safe harbor against being deemed to have knowingly disseminated deceptive synthetic content. Section 7 would let the attorney general issue civil investigative demands, sue, and seek civil penalties of $15,000 per violation, with each day counted as a separate violation for covered providers, large online platforms and capture device manufacturers. The attorney general could also recover reasonable attorney costs and fees and make rules. Section 8 would make people civilly liable for knowingly or recklessly disseminating deceptive synthetic content likely to harm the depicted person. It would create a private right of action with the greater of $1,000 per view or interaction, or actual damages, plus attorney fees. It lists exceptions (such as law enforcement, legal proceedings, satire and news reporting) and allows privacy protections for plaintiffs. Section 9 exempts products offering exclusively non-user-generated video game, television, streaming, movie or interactive experiences. Section 10 would add one year to the basic sentence of a noncapital felony when a separate finding shows generative AI was used to assist in committing it. That year would be non-suspendable.
Status history
Current status as of 2026-01-22
Introduced
2026-01-22
observed 2026-08-25
Impact areas
- Enterprise Adoption
- AI Policy
- Privacy & Data Protection
- Safety & Harms