Including Artificial Intelligence
AI systems tied to infrastructure get new safety rules
Under the Ohio proposal, operators would need a risk-management policy before or soon after deployment. The requirement follows recognized standards and does not cover tools limited to routine or cybersecurity tasks.
The bill would set a new planning duty for anyone running an AI system that helps control a critical infrastructure facility. Ohio would also bar state and local governments from restricting computational tools unless the limit is narrowly tailored to a compelling public need.
- State and local restrictions on computing tools would face a much tighter test.
- AI systems tied to critical infrastructure would need risk-management policies.
- The bill still leaves room for fraud, deepfakes, nuisance and government-use rules.
- A proposal in Ohio would make it harder for state and local governments to wall off new computing tools, including artificial intelligence (AI)
- The Right to Compute Act says restrictions on lawful use, development, deployment or possession would have to clear a very high legal bar
A proposal in Ohio would make it harder for state and local governments to wall off new computing tools, including artificial intelligence (AI). The Right to Compute Act says restrictions on lawful use, development, deployment or possession would have to clear a very high legal bar.
The practical effect is to protect the use of computing power unless officials can show a compelling public reason that cannot be achieved without the restriction. At the same time, the bill would require risk-management policies for AI systems that in whole or in part control critical infrastructure facilities.
The lines the bill draws
The measure defines computational resources broadly, covering systems, software, networks, devices, infrastructure, algorithms, cryptography and AI systems that process or move data. That matters because the bill is not aimed at one narrow product or platform.
It also spells out the kinds of public problems the state could still act on: keeping critical infrastructure reliable, fighting fraud and deceptive practices, limiting harmful AI-generated deepfakes made without consent, dealing with nuisances tied to computing infrastructure, and setting rules for AI use by state and local employees.