United States

• 2026-10-06:The Colorado Attorney Generalreleased interim draft ADMT and conversational-AI rules under SB 26-189; comments close 2026-10-26.
• 2026-10-02:EO 14434was published in the Federal Register.Watch:the proposed legislative definition due about 2026-11-28, which could eventually affect statutory AI definitions (an inference; nothing in the sources says the outcome is decided).
• 2026-03-25:The White House issued a non-binding legislative-policy framework recommending preemption of certain state AI laws. No federal moratorium on state AI laws has been enacted.

No new executive order, DOJ filing or FTC primary announcement was located for 2026-10-07 to 2026-10-10; that is a search limitation, not proof of absence.

There is no comprehensive federal AI statute. The main instruments are:

• Executive Order 14365(2025-12-11).Adopts a posture of federal preemption toward state AI laws and led to the DOJ AI Litigation Task Force, set up by memorandum on 2026-01-09. A related Commerce review is recorded as overdue.
• Executive Order 14434, “Inaugurating the Era of Super Intelligence” (signed 2026-09-29; 91 FR 63129, 2026-10-02).Directs agencies to use the terms “Super Intelligence” or “SI” in non-statutory documents and requires proposed legislative language for a statutory definition within 60 days (about 2026-11-28).
• NIST AI Risk Management Framework(voluntary).No change to the framework itself was found in the latest run.
• Texas Responsible AI Governance Act(TRAIGA, HB 149).In force since 2026-01-01.
• California SB 53, the Transparency in Frontier Artificial Intelligence Act.In force since 2026-01-01.
• Colorado SB 26-189(ADMT Act).Repeals and replaces the earlier Colorado AI Act (SB 24-205) with effect from 2027-01-01.

At federal level, executive orders create no private rights and no new obligations for businesses on their own; Section 5 of the Federal Trade Commission Act (unfair or deceptive practices) remains the enforcement floor.

The state layer is where binding AI rules sit. Texas TRAIGA covers AI development and deployment, bans certain uses, and is enforced exclusively by the Attorney General (no private right of action). California SB 53 applies to large frontier developers, meaning models trained with more than 10^26 floating-point operations, and requires published frontier AI frameworks, transparency reports and safety protocols. Colorado’s replacement statute takes a narrower disclosure-and-rights approach to automated decision-making technology (ADMT).

Obligations must be mapped state by state. Texas TRAIGAcarries civil penalties of up to US$200,000 per violation, with a 60-day cure period.

• Colorado.The Attorney Generalreleased interim draft rules on ADMT and conversational AI on 2026-10-06; comments are open until 2026-10-26.
• Litigation.The DOJ moved to interveneon 2026-04-24 inxAI v. Weiser(D. Colo.) against the earlier Colorado AI Act. Enforcement is reportedly suspended during the litigation; the earlier recorded stay date of 2026-04-27 is not supported by the latest lane report and is unverified.
• FTC.Press reports (2026-09-30)of an FTC probe of major AI developers, and of an “Operation AI Comply” settlement, are unverified: no FTC primary document was found.

This is research assistance, not legal advice; verify before reliance.