Colorado Rewrites Its Landmark AI Law, Scaling Back Disclosure Mandates and Pushing Enforcement to 2027

What changed from the original law

Colorado's 2024 AI Act required companies, governments, and other organisations developing or deploying AI to disclose how their systems help make consequential decisions — such as those affecting hiring, loans, or housing.³⁴ Senate Bill 189 replaces that disclosure requirement with a simpler obligation: notify consumers when AI is being used to make a consequential decision about them, and give them an opportunity to appeal.²³

The original Act had been set to come into effect on 30 June 2026. SB 189 pushes that start date back to January 2027.⁴⁵

Bipartisan support

The bill passed with broad bipartisan backing — 57-6 in the House and 34-1 in the Senate.² The measure now heads to Governor Jared Polis, who has said he will sign it into law.³

Why Colorado pulled back

The Colorado Sun described the outcome as the end of a fierce two-year fight over AI regulation, concluding with a watered-down law.¹ Critics of the original legislation argued that extensive disclosure requirements placed undue burdens on businesses and risked driving AI companies out of the state.⁷ Advocates for stronger rules argued the rewrite weakened consumer protections at a critical moment in AI governance.⁶

National context

Colorado's rewrite comes as states across the US grapple with how aggressively to regulate AI, amid pressure from the federal government to stand down.⁸ The White House released a National Policy Framework for AI in March 2026 that called on Congress to preempt state AI laws it deemed to impose undue burdens on innovation.⁸⁹

California's Transparency in Frontier AI Act remains one of the most stringent state-level frameworks, while New York's Governor signed amendments to the state's RAISE Act in March 2026, shifting it toward a transparency and reporting-based model.¹⁰