The Michigan House is considering heavy-handed bills that would impose burdensome requirements on artificial intelligence systems and likely drive AI development out of the state. Several other states are pursuing better approaches, dubbed “Right to Compute” laws. These states are not taking the punitive and one-sided approach Michigan is considering. They instead seek a more balanced one that promotes innovation and puts up guardrails against the most likely dangers of AI development.
“Computing power is not a mere convenience,” technology analyst Taylor Barkley recently explained to an Ohio House committee. “It is the foundation of technological progress, economic growth, and civic participation.” And government’s role should be to make it possible for free people to pursue that progress. “Computational freedom is not a privilege to be granted by the government,” Barkley continued, “but a natural extension of rights we already possess that should be protected by the government.”
Unfortunately, laws proposed in the Michigan Legislature treat new computing applications as inherently suspect and requiring intrusive government oversight. House Bill 4667 creates new felony offenses and mandatory prison sentences for the criminal use, development, or distribution of AI systems. House Bill 4668 would require AI developers in Michigan to publicly disclose their safety protocols, conduct regular risk assessments and comply with third-party audits.
As we pointed out previously, these bills ignore the value of technological innovation. Instead, they focus solely on imposing regulatory restrictions, meaning that lawyers seeking new occasions for bringing lawsuits would have plenty of work. The broad definition of AI contained in the bills could lead to severe unintended consequences. For example, small business owners accused of minor regulatory violations involving how they manage their customer data in Excel spreadsheets could be charged with a felony that carries a mandatory eight-year minimum prison sentence.
These proposed laws for Michigan fail to consider the importance of promoting technological innovation. Fortunately, other states offer a model for how to strike a balance between encouraging economic development and promoting safeguards for AI technology through risk management policies.
The first Right to Compute law was enacted by Montana in April 2025. The Montana law affirms the right to own and use computational technology. It also requires that any new regulations for AI technology be “demonstrably necessary and narrowly tailored to fulfill a compelling government interest in public health or safety.”
At the same time, the Montana law itself imposes two important AI safeguards. First, it mandates that companies create “a shutdown mechanism allowing reversion to human control within a reasonable time.” Second, Montana requires that companies conduct “an annual risk management review, including fallback procedures and mitigation plans.”
Thus, AI-intensive industries and developers know that when they operate in Montana, the state will protect their rights to be innovative while also requiring them to follow a short list of critical requirements for operating safely. Idaho and New Hampshire are also considering their own versions of the law.
Right to Compute is now coming to the Great Lakes region, with Ohio moving toward its own version. As Greg Lawson of The Buckeye Institute recently testified, the proposed Ohio law establishes “clear regulatory guardrails to properly balance innovation and responsible oversight of emerging technology.” It “provides regulatory certainty for the AI sector by preventing local governments from imposing a patchwork of contradictory rules that are expensive and difficult to follow.”
As Right to Compute laws spread, states that adopt them will gain a critical competitive advantage over those that don’t. Manufacturing, energy, and health care are among the industries that will rely the most on artificial intelligence. As the AI arms race comes to universities, the institutions that are hampered by state restrictions on AI use will be at a disadvantage in attracting researchers and in preparing students for high-tech careers.
The Michigan Legislature has a choice to make. It can follow the heavy-handed regulatory approach of HB 4667 and 4668, which will lead to Michigan trailing the rest of the country in technological progress. Its businesses and most creative people will be encouraged to move to more welcoming states or countries. Or Michigan can take a leadership position in AI systems and applications by adopting its own
