The new law replaces strict testing and reporting rules with disclosure steps after businesses and federal officials challenged the original law.

Colorado lawmakers passed a revised version of the Consumer Protections for Artificial Intelligence Act (formerly Bill 24-205), which replaces several testing and reporting steps with disclosure requirements.

The new bill, Senate Bill 189, applies to systems that help make decisions about jobs, loans, housing, healthcare, and insurance. It requires companies to inform people when they use an automated system to make a decision. It also gives individuals the right to request a human review and correct incorrect data used in decision-making.

The original 2024 law required companies using “high-risk” systems to conduct impact assessments to identify discriminatory outcomes, create risk management plans, conduct annual AI system reviews, and document how systems worked and made decisions.

Businesses pushed back, claiming the law would require changes to how systems produce results and increase compliance costs. They also said the rules could limit how systems are designed and deployed, stifling AI innovation.

Governor Jared Polis responded by forming a coalition of companies, legal advisors, and policy staff to review how companies would carry out testing, reporting, and disclosure requirements. The group produced a revised proposal after months of negotiation and multiple drafts.

xAi’s lawsuit challenging the law’s constitutional viability only accelerated the process. The U.S. Department of Justice joined the case and raised similar legal arguments.

Lawmakers used the working group’s proposal to rewrite the law and pass the new bill with broad support.

According to The Colorado Sun, Senate Majority Leader Robert Rodriguez described the result as a compromise. “Everybody lost and everybody won,” he said.

The bill now goes to Governor Polis, who is expected to sign it. If ratified, the law goes into effect January 1, 2027.