HB 5579, the "Responsible Artificial Intelligence Security for Employees Act," prohibits most employers from using automated decisions tools (like AI for hiring) or electronic monitoring tools (like tracking software) for employment decisions affecting employees or job applicants. Employers may only use these tools for specific, limited purposes, such as screening large applicant volumes for job skills or monitoring essential work functions, safety, or compliance. The bill requires employers to obtain written consent from workers, provide clear notice of monitoring, ensure data accuracy, and use tools in the least invasive way possible. It directly affects all Michigan employers and their employees/applicants, creating new rules for technology use in the workplace.
HB 4667 prohibits using artificial intelligence (AI) to commit crimes in Michigan. It makes it a felony to develop, deploy, or distribute AI systems with the intent to commit other crimes (8-year sentence), or to use AI to further such crimes (8-year sentence added to the underlying crime). The law also targets those who design or distribute AI systems knowing they will be used for criminal purposes (4-year sentence). It excludes law enforcement officers using AI in official duties and defines "AI system" broadly to include generative tools like chatbots, but excludes basic search engines or plug-ins. The bill would require mandatory prison terms with no parole or probation for violations.
HB 4938, the "Anticorruption of Public Morals Act," prohibits the online distribution of specific material defined as "prohibited" under the bill. It directly affects internet service providers, social media platforms, websites, and commercial entities by requiring them to block access to material including most pornographic content (depicting specific sexual acts, AI-generated sexual content, or content involving gender expression that "disconnects biology and gender"), with exceptions for scientific/medical research and academic content. Key mechanisms include mandatory filtering systems for ISPs, requirements for platforms to implement AI and human content moderation with 2-day response times for flagged content, and annual transparency reports. Violations carry criminal penalties (up to 25 years in prison or $125,000 fines) or civil fines of up to $500,000 per violation, enforced by a new Attorney General's division.
This bill prohibits Michigan's medical assistance program (Medicaid) from using artificial intelligence tools to deny, modify, or delay claims. It directly affects the state Department of Health and Human Services and private health plans contracted to administer benefits. The law explicitly bans AI-driven reviews as the sole basis for claim decisions, requiring human review instead. This change aims to prevent automated systems from making final determinations on healthcare coverage eligibility or payments.
HB 4536 prohibits health insurers in Michigan from denying, modifying, or delaying insurance claims based solely on artificial intelligence reviews. It directly affects health insurers operating in the state and the policyholders whose claims are processed by these insurers. The bill amends Michigan's Insurance Code (MCL 500.100-500.8302) by adding Section 3406ss, which requires insurers to use human review for claim decisions instead of relying exclusively on AI tools. This creates a clear policy change: insurers must not use AI as the sole basis for adverse claim actions.
