USA:Colorado Artificial Intelligence Act (CAIA)

1.Overview

Colorado became the first U.S. state to  enact the Colorado Artificial Intelligence Act(CAIA) via Senate,in May 2024.The CAIA definition of  an AI system is“any machine-based system that, for some direct or indirect purpose, gathers from the inputs received by the system how to produce outputs such as content, decisions, predictions, or recommendations, that can affect physical or virtual environments.2.Scope

The CAIA provides an exemption for high-risk AI Deployers if they are small to medium-sized enterprises (SMEs) employing 50 or fewer full-time employees and meet certain conditions, they do not need to maintain a risk management program, conduct an impact assessment, or create a public statement. The SMEs are still subject to a duty of care and must provide the relevant consumer notices.

3.Key provisions

• Prohibited Act:The CAIA prohibits the development or deployment of high-risk AI systems without  reasonable care to mitigate algorithmic discrimination.Developers and deployers are also mandated  to perform impact assessments and keep documentation in evidence of compliance.
• Consent Requirements:Although the CAIA does not specifically require consent for AI deployment, it puts forward transparency and accountability,by compelling open disclosures on AI use,particularly in high-risk situations.​
• Platform Responsibilities: Platforms that monitor high-risk AI systems are required to have governance programs in place, perform impact assessments, and issue transparency reports to the monitor. ​

4.Penalties & Enforcement

• Penalties:Any  CAIA violation is a deceptive trade practice that is subject to significant civil penalties under the Colorado Consumer Protection Act.  Section 6-1-112 of the Colorado Consumer Protection Act now imposes a civil penalty of up to $20,000 per violation, and up to $50,000 per violation if a deceptive trade practice is done against a senior citizen who is  over age 60.
• Enforcement Authority:The CAIA grants the Colorado AG sole enforcement authority.It is under the Colorado Attorney General, with a rebuttable presumption of compliance for those meeting documentation and governance standards.

5. Notable Cases or Precedents

Since the CAIA is scheduled to come into effect on February 1, 2026, there have been no significant cases or legal precedents so far.​

6. Comparison to Global Standards

The Colorado AI Act is primarily concerned with high-risk AI systems defined as any AI system that makes or is a significant factor in making a consequential decision. This definition of a high-risk system is slightly different from that used by the EU AI Act, which is more specifically concerned with determining  systems that cause serious harmful effects on health, safety, and essential rights.

7. Practical implications 

Under the law, deployers of high-risk AI systems are required to use reasonable care to prevent algorithmic discrimination. As with developers, deployers are presumed to have exercised reasonable care as long as they implement a risk management policy and program for high-risk AI systems are required to set up governance programs and file transparency to ensure they are complying with the CAIA.

8.Future outlook

The CAIA established a  precedent for regulating AI in the United States, and its enactment could be used to sway other states into passing similar bills.With technology advancement of AI technology,there will come a time when more changes and regulations are made to cope with upcoming challenges as well as to ensure ethical AI implementations.