WeVote
Log In

Bill

Bill

HB 25B-1008

Consumer Protections for Artificial Intelligence Interactions

2025 First Extraordinary Session Introduced by Judy Amabile and 3 co-sponsors

Colorado AI consumer protections require disclosures, risk management docs, anti-discrimination rules, and public-sector contract safeguards, enforceable under state law.

House Third Reading Laid Over Daily - No Amendments
0
WeVote Research Nonpartisan
Bill Summary · HB 25B-1008

Summary — HB 25B‑1008: Consumer Protections for Artificial Intelligence Interactions

Status: Introduced Aug 21, 2025; House Third Reading laid over daily (no amendments) (as of Aug 25, 2025). Legislative Council’s final fiscal note indicates the bill was deemed lost in the House on Aug 26, 2025; impacts therefore would not take effect.

Main purpose

HB 25B‑1008 is intended to create and clarify consumer protections for interactions with AI systems, to define enforcement pathways, and to require public‑sector contracting safeguards for AI. During committee consideration the bill was amended in different ways — the principal amendments either (a) delayed the effective date of provisions in Senate Bill 24‑205 to October 1, 2026, or (b) (in an earlier version) repealed SB 24‑205 and replaced it with a new statutory consumer‑protection framework for AI interactions.

Key provisions (across versions)

  • Disclosure requirements

    • Developers/deployers must disclose when a consumer is interacting with an AI system (not a human) in certain circumstances.
    • A narrower category of “required‑disclosure AI systems” (e.g., systems used for employment decisions, lending/credit, essential government services, healthcare, housing, insurance, legal services, education) must provide clear, conspicuous pre‑interaction disclosure and certain identifying information (developer name, product/trade name, deployer contact).
    • Exceptions for obvious cases or emergency communications.

  • Anti‑discrimination and documentation obligations

    • Developers and deployers of “high‑risk” AI systems must maintain documentation (model cards, dataset cards, impact assessments), adopt risk‑management policies, and perform regular impact assessments and reviews.
    • Use or deployment of AI that results in discriminatory outcomes may create liability under the Colorado Anti‑discrimination Act.

  • Enforcement and remedies

    • Violations can be enforced as deceptive trade practices under the Colorado Consumer Protection Act; the Attorney General has enforcement authority (and may adopt implementing rules).
    • Individuals may file discrimination complaints with the Colorado Civil Rights Division where applicable.

  • Public contracts and indemnity

    • Contracts by public schools, state agencies, and public entities must not waive consumer/anti‑discrimination protections.
    • Contracts acquiring AI services must include developer/deployer certifications of compliance and indemnification/assumption of liability for noncompliance.

Who is affected

  • Consumers interacting with AI in consumer contexts listed above.
  • Developers and deployers of AI systems operating in Colorado.
  • Public entities and schools that purchase or deploy AI systems.
  • State agencies (Office of Information Technology, Department of Law/Attorney General, Judicial Department, Department of Regulatory Agencies) responsible for implementation, oversight, and enforcement.

Fiscal and administrative impacts

  • Legislative Council and JBC analyses generally projected minimal net state revenue effect and modest or shifting workload for state agencies (AG, OIT, Judicial). No large statewide appropriation was required in the versions analyzed.
  • One draft projected modest reductions in Judicial Department expenditures; a JBC amendment shifted $27,786 GF within the Judicial Department budget to reflect contract/fund changes.
  • The final Legislative Council fiscal note states that because the bill was deemed lost in the House on Aug 26, 2025, those impacts would not take effect.

Timing / effective dates (as proposed)

  • Some AI disclosure and consumer protections were drafted to take effect Jan 1, 2027 (disclosure/consumer enforcement provisions).
  • Provisions in SB 24‑205 (related to algorithmic discrimination/risk management) were delayed in amended versions from Feb 1, 2026 to Oct 1, 2026.
  • The bill generally takes effect 90 days after adjournment sine die, unless otherwise specified or a referendum is filed.

Note: HB 25B‑1008 underwent multiple amendments in committee, and the versions vary materially (delay of SB 24‑205 vs. repeal/reenact with new standalone protections). The summary above highlights the core themes and the major differences reflected in legislative and fiscal documents.

Compiled from official sources — confirm details with the bill’s official record.

Sign in to ask a question.