With less than two weeks left in the 2026 legislative session, Colorado Senate Majority Leader Robert Rodriguez on Friday introduced one of the most anticipated bills of the year, setting the stage for significant reforms to the state’s artificial intelligence regulations.
Senate Bill 189 is the culmination of a two-year effort to amend the 2024 law, which would be the most comprehensive AI regulation in the country, but everyone from technology leaders to school districts deemed it too burdensome. Both the task force and a governor-appointed task force are working to come up with a solution, and the task force proposed a detailed framework six weeks ago, much of which is incorporated into the new bill, which will hold its first committee hearing early next week.
SB 189 specifically seeks to regulate automated decision-making technologies used to make consequential decisions regarding individuals and to ensure that the results do not involve algorithmic discrimination. The law defines major decisions as those that affect education, employment, housing, financial services, insurance, health care, and essential government services, affecting everything from job application considerations to important decisions about medical procedures.
In an interview on the Colorado Chamber of Commerce Office Hours podcast, Rodriguez said the tone and focus of the new bill is different from Senate Bill 24-205, which a federal court recently halted enforcement of. The law is scheduled to go into effect on June 30, but Attorney General Phil Weiser has delayed rulemaking while lawmakers are currently trying to come up with an alternative, represented by SB 189.
How will this change existing AI laws?
Colorado Senate Majority Leader Robert Rodriguez spoke about the AI bill on the Colorado Chamber of Commerce Office Hours podcast.
The 2024 law required developers and adopters to comprehensively assess the discrimination risks of AI systems, and required users and the attorney general’s office to uncover potential algorithmic problems. But SB 189 is focused on making sure people know that AI is helping them make important decisions about them and allowing the system to correct the information it has about them if there is an unfavorable decision.
“This is not so much a comprehensive disclosure bill; it’s more of a notification bill,” Rodriguez said. “The requirements are much less stringent, but I think we can still achieve some of the basic goals.
“It’s not everything I want there, it’s not everything consumer groups want it to be there, it’s not everything the companies want it to be there … and sometimes you’re in a good position when everyone’s not happy,” the Denver Democrat continued. “I didn’t get everything I wanted, but I’m relieved that others didn’t either.”
This sense of effective compromise is shared by Liz Peets, Comcast’s vice president of government and community affairs, who was deeply involved in the negotiations.
“The product of many compromises”
He said all industries, including small and medium-sized businesses that may use AI for purposes other than resume screening, will need to determine how the bill’s requirements will change the way they operate. But unlike some of the mandates in the 2024 law, these regulations appear to be actionable for businesses of all sizes, she said.
Liz Peets, Comcast’s vice president of government and community affairs, talks about the AI bill on the Colorado Chamber of Commerce Office Hours podcast.
“The good thing about this bill is that more companies are in a position to say, ‘This is slightly better than 205,'” Peets said. “Unraveling this responsibility was the most difficult part because the positions of consumers and business groups were so far apart, and this bill was the product of many compromises.”
After signing the 2024 law with reservations, Polis formed a task force to try to resolve the issue during the 2025 regular and special legislative session, but no agreement could be reached. He then organized a working group that met behind closed doors and held lengthy discussions until an agreement was reached.
Liability was one of the biggest issues during last year’s special legislative session.
SB 189 repeals the joint and several liability proposal that undermined the special session’s efforts and replaces it with a provision that allocates fault between the developer and adopter of an AI system based on the intended use of the system in the resulting decision. The Attorney General’s Office is solely responsible for enforcing laws under Colorado’s anti-discrimination law, and the bill does not create any new private rights of action.
Responsibilities of AI developers and adopters
Developers of AI systems need to provide adopters with an explanation of what their technology will be used for, as this is a nascent but growing industry where neither police nor business leaders want to fear overregulation. Adopters (any business, nonprofit, or government organization that uses AI to make consequential decisions) must notify consumers of the existence of automated decision-making technology at the time they interact with covered AI systems.
Within 30 days after a resulting decision is made, implementers must inform consumers of the AI system’s role in that decision and provide consumers with the right to meaningful human review of adverse decisions and the right to rectify inaccurate personal data that may have been used in the decision. Such requirements would be far less onerous than those outlined in the 2024 law, which some critics said could require allowing individual consumers to appeal decisions that are unfavorable to both AI developers and adopters.
Finally, while the new bill would give deployers and developers the right to correct identified violations at the request of the task force, Rodriguez added a three-year deadline to that provision to ensure that affected parties do not have an indefinite get-out-of-jail-free card, he said. SB 189 requires the Attorney General to complete rulemaking on certain provisions of the law by the end of this year, including how to handle adverse income notices, in time for the law to go into effect on January 1st.
Less strict, but no less effective?
Colorado Senate Majority Leader Robert Rodriguez speaks on the 2024 AI bill on the Senate floor.
Rodriguez, a leader in the national debate on AI regulation, acknowledged that he initially wanted a stricter framework, including requirements such as testing algorithm developers for potential discrimination. But he believes the bill’s remaining liability requirements will serve as an incentive for these companies to ensure their products do not cause harm before they go to market.
The U.S. Department of Justice is also joining a lawsuit filed by Elon Musk’s xAI alleging that the 2024 law violates developer rights, but Rodriguez said he did not structure SB189 to address the federal government’s concerns. He believes the lawsuit is without merit and that courts will find that states, and perhaps eventually the federal government, have the right to regulate artificial intelligence to protect residents from harm.
“If you look at the polls, most people think we don’t regulate AI enough…I think technology should be built on trust,” said Rodriguez, who brought on Senate President James Coleman as the bill’s lead co-sponsor. “Is this a perfect bill? No. But it’s never going to be perfect, and technology is going to change a lot.”
SB189 received an initial vote of confidence Friday from the People’s Alliance for Responsible Technology, a coalition of labor and citizen groups formed after a working group submitted recommendations calling for transparency and accountability from Big Tech companies. The group issued a statement saying it was “cautiously optimistic” about the bill because it “provides a path to holding developers and companies that use AI accountable” for their resulting decisions.
Other AI bills continue to be debated
Colorado Congressmen Javier Mabry and Gretchen Rydin debated a bill on the House floor last month that would regulate AI in psychotherapy.
The consumer concerns that led to the development of SB 189 have led to other AI-focused bills this Congress, including measures to regulate their use in health care and psychotherapy, and to put more guardrails around the increasingly popular chatbots. All three bills are still pending in the General Assembly.
Peets said individual businesses should now consider the bill to understand how it will affect them and whether those regulations are acceptable. The biggest impact for her company will likely involve the use of AI in recruiting, but everyone from home sellers to insurance executives to loan officers will be affected differently by SB 189 and need to understand what they must do in response.
But in general, she feels the proposal would be easier for small businesses to comply with than current law, especially in that it would not require them to conduct similarly extensive evaluations of potential flaws in AI systems. And now, after almost four years of working on this issue, going back to the original two-year process of developing SB 24-205, she feels this new bill has fallen into place.
“What the business community is looking at is, ‘Are the disclosures something we can comply with?'” she says. “Is it perfect? No. But I think everyone feels like they got a little bit out of this process.”
