Colorado Lawmakers Advance Revised AI Law After Years of Debate
Key Takeaways
- Colorado lawmakers introduced SB 189 this week, a comprehensive rewrite of the state's 2024 AI law that scales back many of the original requirements after industry pushback and legal challenges. The Colorado AI law amendments shift focus from regulating "high-risk AI systems" to targeting "covered automated decision-making technology" that materially influences consequential decisions in areas like employment, housing, and healthcare.
- The new Colorado SB 189 AI regulation removes requirements for deployers to conduct impact assessments and eliminates the vague "algorithmic discrimination" standard, instead relying on existing anti-discrimination laws to address bias concerns. Developers must still disclose key information about their systems to deployers, including intended uses, training data, and limitations.
- The automated decision-making technology regulation preserves transparency requirements, mandating that deployers notify consumers when ADMT is used in consequential decisions and provide additional disclosures within 30 days of adverse decisions. The bill passed the Senate 34-1 and has support from the ACLU and many industry groups, though labor organizations argue it weakens worker protections.
- Efforts come as momentum for broad AI regulation has slowed nationwide, with no other state adopting Colorado's original framework and President Trump threatening to withhold federal broadband funds from states with onerous AI laws. If enacted, the revised law would take effect January 1, 2027, potentially offering a framework that balances industry concerns with consumer protections.
- If you're a subscriber, click here for the full edition of this update. Or, click here to learn more about our MultiState.ai+ subscription.
Colorado became the first state in the nation to enact a comprehensive regulatory framework for artificial intelligence when lawmakers passed the AI Act 2024 (CO SB 205) that was set to take effect this year. But despite brief celebration as the first-in-the-nation broadly applied AI law, almost from the moment it was signed, there was broad agreement that the law would need to be revisited. Industry groups pushed back against what they claimed were vague and onerous obligations, and even supporters acknowledged that key provisions would require refinement before taking effect. Efforts to fix the law have proven difficult and politically fraught, but after years of debate, a new AI regulatory bill introduced this week may finally represent a framework that can command a broader consensus.
Last year, Colorado lawmakers attempted to address concerns with the artificial intelligence law but could not reach a consensus, opting instead to delay implementation by pushing the law’s effective date to June 2026. With just two weeks remaining in the current legislative session, the bill’s lead sponsor, Sen. Robert Rodriguez (D), has introduced a full rewrite. The newly introduced bill (CO SB 189) largely tracks the framework released by the governor’s AI policy working group, which we wrote about last month.
If you're a subscriber, click here for the full edition of this update. Or, click here to learn more about our MultiState.ai+ subscription.
