Source: HireRight

Colorado enacted the Colorado Artificial Intelligence Act in 2024, imposing obligations on developers and deployers of "high-risk" AI systems used in consequential decisions including employment. Governor Jared Polis signed the Act but reservations emerged from industry and policy stakeholders about scope, operational complexity, and the risk of outpacing technology. The statute required organizations to adopt risk management programs, conduct impact assessments, and provide detailed disclosures tied to AI influence on decisions. Employers and vendors found the high-risk classification difficult to apply across hiring workflows where multiple tools shape outcomes. Colorado delayed the law’s implementation to June 30, 2026 to create space for reconsideration of scope and structure. An Artificial Intelligence Policy Workgroup convened stakeholders and reached unanimous support for a revised framework.

The replacement bill shifts the regulatory focus from defining "high-risk AI systems" to whether technology materially influences a consequential decision, capturing a broader range of automated processes and emphasizing a tool's role in decision-making rather than its label. The proposal removes pre-deployment and ongoing impact assessment mandates and drops the formal risk management program requirement from the original statute, while emphasizing documentation and the ability to explain how tools are used within hiring workflows. Disclosure changes allow more general notice at interaction and place greater weight on post-decision communication, particularly for adverse outcomes. The bill narrows data correction rights to factual inaccuracies, clarifies allocation of responsibility between developers and deployers, and introduces recordkeeping requirements and a broader definition of covered technology. For employers and background-screening vendors the change reframes compliance from building internal governance programs to maintaining visibility into decision processes.