SB 1012

As substituted on 3/9/2026: - Clarifies AI systems to be non-sentient and cannot be granted personhood. Prohibits AI from marrying, serving in corporate roles, or owning property. - Clarifies owners and users who direct an AI are liable for harms it causes. - Requires owners and operators to maintain documented oversight, supervision, and risk-controlled measures for any AI system whose outputs or recommendations could reasonably be expected to impact human welfare, property, or public safety. - Requires developers, manufacturers, and owners of AI systems to prioritize safety mechanisms to prevent or mitigate risk of direct harm. Regular risk assessments may be required. - Allows courts to pierce the corporate veil when AI systems cause significant harm. - Requires owners, operators, and developers of AI systems to notify the attorney general for incidents creating a substantial risk of harm. - Requires a disclaimer on political advertisements that use generative AI to create content intended to injure a candidate or deceive voters. - Creates the criminal offense of disclosing or threatening to disclose a deepfake, classifying the act as a class E felony, or a class B felony if the deepfake is of an intimate nature. - The original version of the bill related to political deepfakes. The 2/18/2026 substitution added provisions relating to personhood and liability.