Colorado lawmakers are moving to substantially rewrite the state’s landmark 2024 artificial intelligence law, advancing a compromise bill that would both scale back and clarify some of the nation’s most stringent AI compliance requirements.
Senate Bill 189, introduced in the final days of the 2026 legislative session, would effectively repeal and replace the original law with a framework negotiated by a governor-appointed task force and a broad stakeholder working group. The measure reflects what Senate Majority Leader Robert Rodriguez described as “98%” alignment with the task force proposal, albeit with targeted revisions intended to make the regime easier to administer.
The 2024 law was widely viewed as the most comprehensive state-level AI statute in the U.S., imposing obligations around algorithmic discrimination, risk management programs, impact assessments, and disclosures. But industry groups and some policymakers argued it was overly complex and impractical to implement.
According to Colorado Politics, SB 189 seeks to address those concerns by restructuring compliance expectations while preserving core consumer protections. Among the most notable changes is a delayed implementation timeline, pushing enforcement to later in 2026 to give the Colorado attorney general additional time to complete rulemaking.
The bill also revises enforcement mechanics under the Colorado Consumer Protection Act, particularly through the introduction of a time-limited “right to cure” provision. Under the proposal, companies would have an opportunity to remedy violations before facing enforcement action, but only for a defined period.
Rodriguez has pushed for a three-year sunset on that provision, arguing it should not be permanent. As drafted, the right to cure would expire on Jan. 1, 2030.
Related: EU Opens Public Feedback Process on Landmark Merger Rule Overhaul
That provision remains contentious. Adam Burrows of Range Ventures described the revised bill as “a massive improvement” over the original law but flagged concerns about how the cure period would apply to future market entrants. “Start-ups that begin in two years… may not know which disclosures apply to them,” he told CP.
Despite the scaling back of certain compliance burdens, the revised framework maintains a strong emphasis on transparency and individual rights in high-stakes automated decision-making.
Gov. Jared Polis underscored that the proposal would still require upfront notice when AI or automated decision-making systems (ADMs) are used in consequential decisions. “Coloradans will receive an up-front notice… and if the decision is adverse, then they will have access to more information about the decision, an opportunity to correct wrong information, and request that a human review,” Polis said.
Labor and advocacy groups have cautiously endorsed the compromise. Dennis Dougherty of the Colorado AFL-CIO said the coalition he represents is “cautiously optimistic,” adding that the bill “provides a path to hold developers and businesses using AI accountable when the technology makes consequential decisions.”
Business stakeholders appear to view the measure as a workable middle ground. Loren Furman, head of the Colorado Chamber of Commerce, said the proposal “strikes a reasonable balance,” reflecting the working group’s consensus even as the chamber continues to seek amendments.
The legislative rewrite is unfolding against a high-stakes legal challenge that could upend the law entirely. Elon Musk’s AI company xAI filed suit in April seeking to block the 2024 statute, arguing it is unconstitutional and requesting a preliminary injunction.
The U.S. Department of Justice has since intervened in the case in support of xAI—an unusual move that signals the federal government’s growing willingness to challenge state-level AI regulation, according to a government contractor compliance update by Proskauer.
In its filings, the government argues the Colorado law violates the Equal Protection Clause by effectively requiring developers to consider protected characteristics such as race or sex in order to mitigate algorithmic bias. The intervention aligns with broader Trump administration policy, which has emphasized removing regulatory barriers to AI development
A federal judge has already issued a temporary stay blocking enforcement of the law while the court considers xAI’s request for an injunction.
For now, companies face a moving target. While lawmakers work to finalize SB 189 and potentially soften the original statute, the litigation introduces the possibility that the law could be delayed or invalidated altogether.
The result is a fluid regulatory landscape in which Colorado continues to serve as a testing ground, not only for AI governance frameworks, but also for the emerging conflict between state-level innovation policy and federal efforts to assert primacy over AI regulation.
