Status: repealed

Senate Bill 24-205 — the first comprehensive US state AI law — was repealed by Governor Jared Polis on May 14, 2026, before its June 30, 2026 effective date. The repeal was paired with the signing of SB 26-189, a substantially narrower replacement framework that takes effect January 1, 2027.

If you are still working from a tracker or summary that cites “Colorado AI Act — June 2026,” that claim is now obsolete.

How we got here

  1. May 17, 2024 — Polis signs SB 24-205, the first comprehensive US state law targeting algorithmic discrimination in high-risk AI systems (employment, housing, lending, education, healthcare, government services).
  2. April 27, 2026 — A federal magistrate judge grants a joint motion from xAI and the Colorado AG to freeze enforcement. AG Phil Weiser commits to no rulemaking or enforcement until the legislative session concludes.
  3. May 7–9, 2026 — Both chambers of the Colorado legislature pass SB 26-189 (replacement bill) and send it to the Governor.
  4. May 14, 2026 — Polis signs SB 26-189 into law. The original SB 24-205 is repealed in the same action.

What the original SB 24-205 would have required

Documented here as historical reference — the framework has not survived in its original form, but the operational categories influenced both SB 26-189 and other US state proposals:

  • Reasonable care to protect consumers from algorithmic discrimination
  • Annual impact assessments for high-risk AI systems
  • Consumer notice that an AI system was used and an explanation if an adverse decision was made
  • Right to correct data and to appeal to human review where technically feasible
  • Disclosure to the AG of any known algorithmic-discrimination incident within 90 days
  • Exclusive AG enforcement; no private right of action; deceptive-trade-practices framing

What SB 26-189 keeps and drops

Keeps: disclosure to consumers when automated decision-making is used, a right to access information about how the decision was reached, AG enforcement authority.

Drops: the EU-style high-risk classification and conformity assessment regime, the mandatory annual impact-assessment requirement, the proactive reasonable-care duty for developers.

The result is a disclosure-and-rights statute closer in posture to California’s evolving privacy framework than to the EU AI Act. Practitioners who built compliance programs assuming SB 24-205 will retain most of the technical scaffolding (model inventories, impact-assessment templates, incident-disclosure pipelines) — but the legal trigger for those obligations in Colorado has narrowed.

What this means for the US state landscape

With Colorado off the board, the live US state AI map as of May 25, 2026:

  • Illinois HB 3773 — live since Jan 1, 2026 (anti-discrimination + employer notice in employment AI)
  • Texas TRAIGA (HB 149) — live since Jan 1, 2026 (prohibited uses + sandbox + AG enforcement)
  • NYC Local Law 144 — live since July 2023 (AEDT bias audits)
  • Colorado SB 26-189 — effective Jan 1, 2027 (disclosure-and-rights)

The “deadlines are converging” narrative is stronger without Colorado, not weaker — Illinois and Texas are already in force, and the EU AI Act’s Article 50 + GPAI obligations land August 2, 2026.