Colorado Lawmakers Move to Replace Contentious AI Law With New Rules

In brief

• Colorado lawmakers plan to repeal and replace the state’s AI law, SB24-205.
• New rules would regulate AI used in high-stakes decisions like jobs and housing.
• The rewrite follows arguments, including those from Elon Musk, that the original law went too far.

Colorado lawmakers are moving to rewrite one of the most closely watched artificial intelligence laws and attempt to reset rules that govern how AI is used in decisions that affect jobs, housing, and access to services.

The proposal would scrap Colorado’s current AI law, SB24-205, passed in 2024, and replace it with new rules meant to address industry concerns while still protecting consumers.

“In 2024, the general assembly enacted Senate Bill 24-205, which created consumer protections in interactions with artificial intelligence systems,” the bill summary reads.
“The bill repeals and reenacts those provisions with new requirements regarding the use of automated decision-making technology in consequential decisions.”

The new bill, SB26-189, narrows the law to systems used in “consequential decisions,” including employment, education, housing, lending, insurance, healthcare, and government services. It applies to tools that process personal data to generate outputs such as predictions or rankings that influence those outcomes.

The move comes after months of pushback from the AI industry over Colorado’s 2024 law, which requires companies to check for and reduce bias in decisions like hiring, lending, and housing.

Concerns about the original law centered on how much responsibility it places on companies developing and deploying AI in the state. In April, Elon Musk’s AI company xAI sued the state to block the measure before it could take effect. The lawsuit was joined by the U.S. Department of Justice, which looked to intervene in the case and support xAI.

“I do think they have a valid argument in terms of the burdens that the Colorado policy would place on these companies,” Cody Barela, a partner at Armstrong Teasdale, previously told Decrypt. “The burden on them, in comparison to the delay that it causes in the AI race, might actually be a better argument.”

According to the new bill, developers would be required to provide documentation explaining how their systems work, what data they use, and their known limitations. They would also need to notify companies of material updates.

Companies using these systems would have to notify consumers when AI is involved in a decision, explain adverse outcomes in plain language, allow individuals to access and correct their data, and request a human review. Developers would also be required to provide documentation on how their systems work, the data they use, and their limitations, and to notify companies of significant updates.

If passed, the new bill would take effect Jan. 1, 2027. The new bill comes as lawmakers in other states, including New York and California, are considering similar rules targeting AI systems, and the Donald Trump Administration seeks to override state laws.

Daily Debrief Newsletter