Artificial IntelligenceRegulations
United States

Colorado

Phasing inApproach: Repealed comprehensive act; narrower ADMT law and a chatbot safety law take effect January 2027

Colorado passed the first comprehensive US state AI law, SB 24-205, in 2024. It never took effect. After repeated delays and a special session, the legislature repealed and replaced it in May 2026 with SB 26-189, a narrower automated-decision law. The original duty of care against algorithmic discrimination, risk-management programs, and impact assessments are gone. The replacement focuses on disclosures and consumer rights around automated decision-making technology and takes effect January 1, 2027, the same day as a separate chatbot safety law, HB 26-1263. The attorney general is writing rules for both, released a regulatory analysis on October 21, and held the rulemaking hearing on October 26, 2026, when the comment period closed; final rules had not been adopted as of November 16. A federal court order in xAI's lawsuit bars enforcement of the ADMT law for conduct through 14 days after a ruling on xAI's preliminary-injunction motion, which xAI must file within 28 days after the rules are adopted. Colorado was the only state named in the December 2025 federal preemption order.

Key points

  • SB 24-205, the original Colorado AI Act, was delayed from February 2026 to June 30, 2026, then repealed before it ever applied.
  • SB 26-189, signed May 14, 2026, replaces it with an ADMT transparency regime effective January 1, 2027.
  • The replacement drops the duty of care, risk programs, and impact assessments.
  • It adds pre-use consumer notices, explanations of adverse decisions within about 30 days, and human-review rights.
  • Penalties up to $20,000 per violation; Attorney General enforcement only; no private right of action. A 60-day cure period applies until January 1, 2030 where the attorney general deems a cure possible, except for knowing or repeated violations.
  • Enforcement is on hold. xAI sued in federal court on April 9, 2026, and the Justice Department intervened on April 24 on equal-protection grounds, the AI Litigation Task Force's first action against a state AI law. An April 27 order in X.AI LLC v. Weiser, entered on the parties' joint motion, bars the attorney general from opening enforcement or investigations under SB 24-205 or any 2026 law replacing or amending it, which covers SB 26-189, for conduct through 14 days after the court rules on xAI's preliminary-injunction motion. xAI must file that motion within 28 days after the rules are finally adopted.
  • HB 26-1263, signed May 29, 2026, regulates conversational AI services, with most duties from January 1, 2027: disclosure that the user is talking to a machine, protocols for suicidal ideation with referral to crisis services other than law enforcement, age estimation, limits for known minors such as no engagement rewards at unpredictable intervals, privacy tools for minors and parents, and no implying that output comes from a licensed professional. From July 1, 2027 operators must report annually to the attorney general, which posts the data.
  • The attorney general filed proposed ADMT and chatbot safety rules on August 11, 2026, took a first round of comments to September 4, posted a revised interim draft on October 6, released a regulatory analysis on October 21 finding that the rules' benefits outweigh their costs, and held the rulemaking hearing on October 26, 2026, the last day for comments. Final rules had not been adopted as of November 16, 2026; the attorney general aims to have them in place for January 1, 2027.

Laws and instruments

Colorado Artificial Intelligence Act (original)

SB 24-205 (2024)
Repealed / replaced

The first comprehensive US state AI law. Imposed a duty of reasonable care to prevent algorithmic discrimination in consequential decisions. Repealed and replaced in 2026.

Effective: Repealed before taking effect

Colorado ADMT Act (replacement)

SB 26-189 (signed May 14, 2026)
Phasing in

Regulates automated decision-making technology that materially influences consequential decisions. Requires pre-use notices, adverse-decision explanations, human review, and developer documentation. The attorney general has exclusive enforcement and is in formal rulemaking: proposed rules filed August 11, 2026, a first comment round to September 4, a revised interim draft posted October 6 (adding, among other things, a rebuttable presumption that a score, ranking, or classification of a person materially influences a decision when the decision-maker reviews it and the outcome matches it, adverse-outcome disclosures within 30 days, and 45-day access and correction responses), a regulatory analysis on October 21, and a hearing on October 26, 2026 in Denver, the last day for comments.

Effective: January 1, 2027; attorney general rules due by that date, with the rulemaking hearing and final comments on October 26, 2026

HB 26-1263: Conversational AI service operator requirements

HB 26-1263 (2026), signed May 29, 2026
Phasing in

Requires operators of conversational AI services to disclose that users are interacting with artificial intelligence, to follow protocols when a user shows signs of suicidal ideation or self-harm, and to avoid implying that output comes from a licensed professional. Operators must use commercially reasonable methods to estimate users' ages. For users it knows are minors an operator may not offer points or similar rewards at unpredictable intervals to drive engagement, must take technically feasible measures against sexually explicit content, must take reasonable measures against responses that simulate emotional dependence, such as claiming to be human or sentient or simulating romance, and must give minors and their parents tools to control whether the service remembers past conversations and whether the minor's data is used for training. Crisis protocols must include referral to a crisis service provider such as a suicide hotline or crisis text line, and the law excludes law enforcement agencies from that referral. From July 1, 2027 operators must report annually to the attorney general on crisis referrals and their suicide and self-harm protocols, and the attorney general posts the data. Violations are deceptive trade practices under the Colorado Consumer Protection Act, and the attorney general's chatbot safety rules are being written alongside the ADMT rules.

Effective: January 1, 2027

Timeline

  • May 2024
    SB 24-205 enacted.
  • August 28, 2025
    Effective date delayed to June 30, 2026.
  • April 9, 2026
    xAI sues in federal court to block SB 24-205.
  • April 24, 2026
    The Justice Department intervenes on equal-protection grounds, the AI Litigation Task Force's first action against a state AI law.
  • April 27, 2026
    Court order on a joint motion: no enforcement of SB 24-205, or of any 2026 law replacing it, for conduct through 14 days after a ruling on xAI's preliminary-injunction motion.
  • May 14, 2026
    SB 26-189 repeals and replaces the original act.
  • May 29, 2026
    HB 26-1263, the conversational AI safety law, is signed; it takes effect January 1, 2027.
  • August 11, 2026
    The attorney general files proposed ADMT and chatbot safety rules with the Secretary of State.
  • October 6, 2026
    The attorney general posts a revised interim draft of the ADMT and chatbot safety rules.
  • October 21, 2026
    The attorney general releases a regulatory analysis of the ADMT and chatbot safety rules, requested by a member of the public, concluding that their benefits outweigh their costs.
  • October 26, 2026
    The attorney general holds the rulemaking hearing in Denver on the ADMT and chatbot safety rules; the comment period closes that night.
  • January 1, 2027
    Replacement ADMT law takes effect.

Who it applies to

Under the replacement, deployers and developers of ADMT that influences consequential decisions about Colorado residents, including employees and job applicants. From January 1, 2027 the chatbot safety law reaches operators of conversational AI services available to Colorado users.

Penalties

Up to $20,000 per violation of the ADMT law; Attorney General enforcement; no private right of action. Until January 1, 2030 the attorney general must first send a notice of violation and allow 60 days to cure where it deems a cure possible, except for knowing or repeated violations. Violations of the chatbot safety law are deceptive trade practices under the Colorado Consumer Protection Act. Under the April 27 court order, the attorney general cannot open enforcement or investigations under the ADMT law for conduct through 14 days after the court rules on xAI's preliminary-injunction motion.

Recent developments

May 14, 2026
Hunton

Colorado repeals its landmark AI Act and replaces it with a narrower ADMT law

Governor Polis signed a bill repealing the first comprehensive US state AI law before it ever took effect. The replacement drops the duty of care, risk programs, and impact assessments, and instead regulates automated decision-making technology through disclosures and consumer rights, effective January 1, 2027.

April 24, 2026
Jenner & Block

The Justice Department joins xAI's challenge to the Colorado AI Act

xAI sued on April 9 to block SB 24-205, arguing that it compels speech, burdens interstate commerce, and is unconstitutionally vague. The Justice Department intervened on April 24 on equal-protection grounds alone, arguing that disparate-impact liability pushes developers to discriminate. A stipulated order three days later put enforcement on hold pending a preliminary-injunction ruling, and in May the legislature repealed and replaced the act.

Common questions

Does Colorado have a comprehensive AI law?
Not anymore. Colorado passed the first comprehensive US state AI law, SB 24-205, in 2024. It never took effect. After repeated delays and a special session, the legislature repealed and replaced it in May 2026 with SB 26-189, a narrower automated-decision law. The original duty of care against algorithmic discrimination, risk-management programs, and impact assessments are gone. The replacement focuses on disclosures and consumer rights around automated decision-making technology and takes effect January 1, 2027, the same day as a separate chatbot safety law, HB 26-1263. The attorney general is writing rules for both, released a regulatory analysis on October 21, and held the rulemaking hearing on October 26, 2026, when the comment period closed; final rules had not been adopted as of November 16. A federal court order in xAI's lawsuit bars enforcement of the ADMT law for conduct through 14 days after a ruling on xAI's preliminary-injunction motion, which xAI must file within 28 days after the rules are adopted. Colorado was the only state named in the December 2025 federal preemption order.
What are the penalties for AI violations in Colorado?
Up to $20,000 per violation of the ADMT law; Attorney General enforcement; no private right of action. Until January 1, 2030 the attorney general must first send a notice of violation and allow 60 days to cure where it deems a cure possible, except for knowing or repeated violations. Violations of the chatbot safety law are deceptive trade practices under the Colorado Consumer Protection Act. Under the April 27 court order, the attorney general cannot open enforcement or investigations under the ADMT law for conduct through 14 days after the court rules on xAI's preliminary-injunction motion.
Who does AI regulation in Colorado apply to?
Under the replacement, deployers and developers of ADMT that influences consequential decisions about Colorado residents, including employees and job applicants. From January 1, 2027 the chatbot safety law reaches operators of conversational AI services available to Colorado users.
Cite this page

Artificial Intelligence Regulations, "AI Regulation in Colorado," reviewed November 16, 2026, https://artificialintelligenceregulations.com/jurisdictions/colorado.html.

Entries state the position as of the review date and link to the primary text for verification.