Last updated:
The Colorado AI Act (SB 24-205), the first broad U.S. state AI law, was repealed in May 2026 and replaced by Colorado SB 26-189, which takes effect 1 January 2027 with a narrower focus on automated decision-making technology.
Originally signed in May 2024 and scheduled to take effect 30 June 2026, the Colorado AI Act (SB 24-205) was the first broad U.S. state law placing duty-of-care obligations on developers and deployers of high-risk AI systems. It introduced annual impact assessments and a rebuttable presumption of reasonable care for organizations aligning their governance with NIST AI RMF or ISO 42001, effectively turning voluntary standards into a legal safe harbor. The law was repealed via Senate Bill 26-189 in May 2026 following legal challenges and industry pushback over the breadth of the original "high-risk AI system" framing. The replacement regime, also numbered SB 26-189, takes effect 1 January 2027 with a narrower focus on automated decision-making technology (ADMT) in employment, lending, housing, education, and healthcare contexts. The rebuttable-presumption mechanism in the original act remains an influential structural model now reflected in other state-level AI regulation.
Real world example:
A Colorado-based insurance company that had been preparing its first annual impact assessment under SB 24-205 reassessed its compliance approach after the repeal. The team mapped its in-scope systems to the new SB 26-189 ADMT triggers, narrowed the scope to consequential-decision use cases, and rebuilt its evidence base against the new deployer impact assessment and consumer notice requirements ahead of January 2027.
