Michigan House Judiciary Committee Discusses AI Regulation

News Summary

On September 10, 2025, the Michigan House Judiciary Committee convened to review legislation aimed at regulating artificial intelligence. The primary focus was on House Bill 4668, which seeks to ensure AI safety and accountability. Discussions included input from AI watchdogs and business representatives, as well as concerns about the risks associated with AI technologies. While the committee agrees on the need for regulation, there are worries about the complexity of creating state-level rules. The committee unanimously approved a motion to refer the bill for further consideration.

Lansing, Michigan – The House Judiciary Committee convened on September 10, 2025, to review proposed legislation regulating artificial intelligence (AI), continuing discussions that began earlier in the summer. The bills discussed aim to ensure safety in AI system development as concerns about potential risks associated with the technology escalate.

State Representative Sarah Lightner, the committee chair and a key sponsor of the regulations, moderated the hearing, which included testimony from both AI watchdogs and business representatives. The objective was to gather a diverse array of opinions on strategies to oversee AI system development and to thwart unlawful applications of AI technologies.

The primary focus of the hearing was House Bill 4668, which establishes requirements for AI developers to train their models using broad and diverse data sets. It aims to implement safety and security protocols by January 1, 2026. This legislation intends to address the growing concerns about risks associated with AI, which include the potential development of weapons and the capacity for cyberattacks facilitated by AI systems.

Testimony during the hearing underscored the pressing need for regulation. Specific concerns were raised by Felix De Simone, organizing director of PauseAI US, who noted the current lack of oversight and accountability in AI training processes. He highlighted views from experts in the field, pointing out the unpredictable risks associated with AI. For instance, there are fears that AI could inadvertently assist individuals in criminal activities, encompassing scenarios such as helping novice terrorists in weapon creation.

Despite the urgency in regulating advanced AI technology, some committee members voiced apprehensions regarding the complexity of such regulations. Tyler Diers from TechNet addressed concerns over the requirement for independent third-party audits, as mandated by House Bill 4668. Diers stressed the absence of established standards for auditing AI safety, indicating a potential challenge in implementing these requirements effectively.

While the committee found common ground on the need for regulation, they expressed concerns about the possibility of creating a complicated system of varying rules at the state level that could lead to confusion among developers and companies operating in multiple states. Consequently, suggestions were made to either monitor how other states develop regulations or to establish a dedicated task force focused on comprehensive AI regulation.

In addition to House Bill 4668, the committee also discussed House Bill 4667, which proposes the introduction of three new felonies for individuals who use AI to facilitate criminal acts. This bill seeks to combat issues such as scamming and impersonation through AI-generated profiles. However, the committee did not advance this bill, opting instead to focus on the more pressing regulatory needs of House Bill 4668.

The committee’s motion to refer House Bill 4668 to the House Communications and Technology Committee was unanimously approved, reflecting a shared recognition of the need to establish safety protocols as AI technologies continue to advance at a rapid pace. The landscape of AI is constantly evolving, and the Michigan House Judiciary Committee is actively working to ensure that regulatory measures keep up with these developments to safeguard citizens and prevent misuse of AI technologies.

Deeper Dive: News & Info About This Topic

• Michigan Advance: Business Groups, Watchdogs Debate Guardrails for AI Developers
• Wikipedia: Artificial Intelligence
• Dykema: Will Michigan’s New AI Framework Become the National Standard?
• Google Search: AI regulation
• WTOL: New Laws Crack Down on Intimate Deepfakes
• Encyclopedia Britannica: Technology
• Michigan Advance: AI Safeguard Bills to Prevent Technology from Committing Harm
• Google News: AI technology