•
•
•
•
•
•
•
Colorado, once the first US state to pass comprehensive artificial intelligence legislation modeled partly on the European Union's risk-tiered approach, has significantly rewritten that law after months of delays, industry pushback, and a federal court ruling that blocked enforcement of its predecessor. The revised framework, enacted through SB 26-189, replaces the original Colorado AI Act (SB 24-205) with a narrower set of obligations, according to multiple law firm analyses tracking the legislation this year.
Colorado's original AI Act, signed in 2024, was designed to regulate "high-risk" AI systems used in consequential decisions such as employment, lending, housing, and healthcare, requiring developers and deployers to conduct impact assessments and take steps to prevent algorithmic discrimination. Its effective date was pushed back multiple times amid business-community concern over compliance costs and vague requirements, and enforcement of the original law was ultimately delayed and, according to legal trackers, blocked following federal court action.
Colorado's legislature responded this year by passing SB 26-189, which Governor Jared Polis signed, repealing key portions of the original law and replacing them with a more limited framework. Legal analysts tracking the change describe it as a deliberate pivot away from the EU's broad, horizontal AI-regulation model toward a narrower set of obligations focused on specific harms and higher-risk use cases.
The original Colorado AI Act positioned the state as a first mover on comprehensive AI regulation in the US, at a time when Congress had not passed federal AI legislation and most states relied on narrower, sector-specific rules (such as targeted employment-AI or biometric-data laws). It required companies developing or deploying "high-risk" AI systems to complete algorithmic impact assessments, disclose AI use to consumers in certain contexts, and take reasonable steps to prevent algorithmic discrimination, with enforcement power resting with the Colorado Attorney General.
Business groups argued the law's definitions were too broad and compliance obligations too costly, particularly for companies operating across multiple states with inconsistent AI rules. Momentum for revision built through early 2026, culminating in the legislature passing SB 26-189 as a near-total replacement rather than a minor amendment.
Legal trackers expect continued rulemaking and guidance from Colorado's Attorney General's office as the new framework's specific requirements are implemented. Businesses operating in Colorado are advised by legal analysts to review updated compliance timelines under SB 26-189 rather than continuing to build toward the original SB 24-205 requirements. More broadly, other state legislatures considering AI bills in their next sessions are likely to watch how Colorado's narrower model performs before deciding whether to pursue broad "high-risk AI" frameworks or more targeted rules of their own.
Did Colorado repeal its AI Act entirely?
Not entirely — SB 26-189 replaces significant portions of the original Colorado AI Act (SB 24-205) with a narrower framework rather than eliminating AI regulation in the state altogether.
Why was the original Colorado AI Act delayed and rewritten?
Its effective date was pushed back multiple times amid industry concern over compliance costs and broad definitions, and enforcement was ultimately blocked following federal court action, prompting the legislature to pass a replacement law.
Does this affect companies outside Colorado?
Directly, only companies developing or deploying AI systems that touch Colorado consumers or employees are covered, but the rewrite is being closely watched nationally as a signal for how other states may approach AI legislation.
How does Colorado's approach now compare to the EU AI Act?
The EU AI Act uses a broad, horizontal risk-tiered framework covering most AI applications, while Colorado's revised law takes a narrower approach focused on specific higher-risk use cases, according to legal analysts tracking both frameworks.
Is there a federal AI law in the US yet?
No comprehensive federal AI law has been enacted as of this reporting; AI regulation in the US remains largely a state-by-state patchwork alongside existing sector-specific federal rules.
Colorado's rewrite of its AI Act marks a significant retreat from the state's original ambition to become an early US leader in comprehensive, EU-style AI regulation. With SB 26-189 now in place, businesses and other states alike are watching closely to see whether this narrower model becomes a template for US state AI policy, or whether comprehensive AI regulation efforts resurface once compliance and enforcement questions are better resolved.
