The proposed "Artificial Intelligence Act" seeks to regulate high-risk artificial intelligence (AI) systems in New Mexico, particularly those that make significant decisions impacting consumers. The Act requires developers to provide detailed documentation about their AI systems, including training data, known risks of algorithmic discrimination, and intended uses. Additionally, developers must report any risk incidents promptly and maintain transparency regarding the systems they deploy. Deployers of high-risk AI systems are mandated to implement risk management policies, conduct regular impact assessments, and inform consumers directly before making consequential decisions based on AI outputs.

Key provisions of the Act include the requirement for direct consumer notification about the purpose and nature of AI decisions, as well as the opportunity for consumers to receive explanations, correct data, and appeal adverse decisions, unless safety risks are involved. The bill also emphasizes the need for accessible communication and outlines procedures for reporting algorithmic discrimination to the state department of justice. Certain information, such as trade secrets, is exempt from disclosure, and compliance with federal laws is maintained. The state department of justice is granted enforcement authority, and violations of the Act are classified as unfair practices under the Unfair Practices Act, with an effective date of July 1, 2026, and rules to be established by January 1, 2027.