Artificial intelligence safety and disclosure requirements established, and civil remedies provided.
HF4532, the “Responsible Artificial Intelligence Safety and Education Act” or “RAISE Act,” would create a new set of state requirements for developers of artificial intelligence models in Minnesota. The bill defines key terms such as artificial intelligence, developer, safety incident, and critical harm, and focuses on models that could be used in ways that cause serious physical, mental, or property harm, including harms tied to weapons, criminal conduct, unauthorized access, or misuse of model weights.
Before deploying an AI model, a developer would have to adopt a written safety and security protocol, keep detailed records of that protocol and testing results, publish a redacted version of the protocol, and provide it to the attorney general. Developers would also have to conduct annual reviews, update the protocol as needed, and disclose safety incidents to the attorney general within 72 hours. The bill prohibits deployment of a model that creates an unreasonable risk of critical harm and bars false or misleading statements in required disclosures.
The bill would add new provisions to Minnesota Statutes chapter 325M governing AI developers and model deployment. It would impose affirmative compliance, disclosure, recordkeeping, and risk-mitigation duties on developers of artificial intelligence models, while giving the attorney general access to safety protocols and authority to enforce violations through civil actions. It also creates a private right of action for injured persons, allowing recovery of damages, costs, attorney fees, and equitable relief, and authorizes substantial civil penalties of up to $10 million for a first violation and $30 million for subsequent violations.
Based on the bill text and the absence of recorded committee testimony or votes in the provided materials, the overall posture of the bill appears precautionary and regulatory rather than partisan or controversial in the available record. The measure is framed around public safety, transparency, and accountability for advanced AI systems, suggesting support from lawmakers concerned about emerging technology risks. No formal vote history or hearing transcript is provided to indicate broader agreement or opposition.
The main points of contention likely center on the scope and burden of the requirements imposed on AI developers. Potential concerns include whether the definitions of “critical harm,” “safety incident,” and “unreasonable risk” are too broad or uncertain, whether mandatory publication and attorney general disclosure could expose sensitive security information, and whether the recordkeeping and testing obligations are too costly or difficult to implement. On the other hand, supporters would likely emphasize the need for transparency, early incident reporting, and strong enforcement tools to prevent catastrophic misuse of AI systems.