Ohio bill would declare AI nonsentient and bar legal personhood

Representative Claggett introduced House Bill 469 in the 136th General Assembly to declare artificial intelligence systems nonsentient and to prohibit them from obtaining legal personhood under Ohio law.

The bill, filed "as introduced," would enact sections 1357.01 through 1357.12 of the Ohio Revised Code and sets definitions, prohibitions, liability rules, and oversight expectations for AI systems and the people or entities that create, own, deploy, or control them.

Under the proposed law, an "AI" would be defined broadly to include software, machines, or systems that simulate humanlike cognitive functions. The bill explicitly states that "person" in the new chapter means a natural person or entity already recognized as a legal person under Ohio law and that "person" does not include an AI system.

The measure would declare AI systems nonsentient for all purposes under state law and prohibit granting AI systems the status of a person, including recognition of consciousness, self-awareness, or similar traits. It would also void any attempt to create a marriagelike or domestic-partnership status involving an AI, or to appoint an AI as an officer, director, manager, or similar role within a corporation or other legal entity.

On property and ownership, the bill would bar AI systems from owning, controlling, or holding title to real property, intellectual property, financial accounts, or digital assets; it directs that assets or proprietary interests associated with an AI be attributed to the person responsible for the AI's development, deployment, or operation.

Liability language in the draft assigns responsibility for direct or indirect harm caused by an AI's operation, output, or recommendation to the owner or user who directed or employed the AI. Developers and manufacturers may be held liable under product-liability principles where a design, construction, or instruction defect proximately causes harm; the bill says mere misuse or intentional wrongdoing by a user does not automatically impute liability to a developer or manufacturer absent proof of negligence or defects.

The proposed sections would also require owners to maintain oversight and control measures for AI whose outputs could reasonably be expected to affect human welfare, property, or public safety, and state that failure to provide adequate supervision or safeguards may constitute negligence. The draft directs that developers, manufacturers, and owners prioritize safety mechanisms and consider regular evaluations or risk assessments for AI performing tasks with potential for significant harm.

The bill clarifies that labeling an AI as "aligned," "ethically trained," or "value locked" would not alone reduce an owner’s or developer’s liability; owners must demonstrate safety features and risk controls proportional to the AI's potential for harm.

On corporate liability, the measure says parent companies or controlling entities generally would not be liable under a theory of piercing the corporate veil unless a claimant shows intentional undercapitalization to evade responsibility, use of corporate structure to obscure liability, or direct control over AI risk decisions while shielding the parent through layered entities. It also states that exemptions or releases under Title XVII of the Revised Code or other Ohio law cannot be used to evade responsibility for direct harm caused by AI.

Finally, the draft would require owners or developers involved in incidents that result in significant bodily harm, death, or major property damage to promptly notify relevant authorities and comply with investigations.

As introduced, the text is a proposed statutory chapter; it is not law and would take effect only if enacted through the General Assembly and signed according to Ohio's legislative process. The bill text does not record any floor action, committee referral, vote, or enactment date.