“No human” companies planned in Delaware are drawing a federal oversight inquiry over who would answer for an AI agent’s conduct. US Rep. James Comer, the House Oversight Committee chairman, asked Gov. Matt Meyer for documents about the proposal and its potential safety risks.
Comer wrote to Meyer that the committee is examining the state administration’s effort to create an Artificial Intelligence Company, or AIC. He said the proposal could let human owners avoid responsibility for reckless and harmful actions by the software they appoint to run a business.
No Human at the Helm: The Proposed AI Company Plan
A draft bill would let an AIC pursue any lawful purpose except banking. It says AI agents could manage the company’s affairs and conduct its business. The proposal also calls for a regulatory sandbox lasting 30 months to test the structure.
Under the draft, an AIC would be a separate legal entity able to hold property, take on obligations, and sue or be sued in its own name. A person or entity would fund it and receive profits, while an AI agent would direct daily operations. The member would have to keep the company adequately capitalized, and the AIC would have to log its activities.
The bill says a member would not be liable for an act or omission by the AIC, including conduct taken at its direction. It describes exceptions for fraud, a willful violation of law, or allowing the company to operate outside its stated purpose when the member knew or should have known. Comer argued that those protections could leave business partners, consumers, and the public exposed to harm caused by the controlling AI.
Federal Questions Focus on Liability
Comer asked for records about how Delaware developed the proposal and considered AI safety. He said the committee’s review could inform federal legislation to limit state AI regulatory initiatives. The letter did not announce a bill or establish that the proposal violates existing law.
The proposal’s supporters have described a different aim. In a July 2026 Fortune op-ed, Delaware Secretary of State Charuni Patibanda-Sanchez and Norm Ai founder John Nay argued that legal recognition could make an autonomous system’s conduct visible and attach responsibility and damages to a defined entity. They said Delaware’s sandbox would test agentic commerce under supervision, with capital tied to liability and a way to shut a system down.
A spokesperson for Meyer said the governor’s office had received Comer’s letter and was reviewing it. A spokesperson for Patibanda-Sanchez declined to comment and referred press requests to the governor’s office.
The draft’s liability language makes the inquiry consequential for companies considering AI-directed operations. Grit Daily has reported on a founder’s guilty plea in a major fraud case and on research about fraud risks at venture-backed startups. Those stories offer related context on accountability and business risk, but they do not address Delaware’s proposed law.
Delaware’s proposal followed earlier state work on an AI regulatory sandbox. Meyer signed a legislative resolution in July 2025 to create a framework for testing technologies using agentic artificial intelligence. The Delaware Artificial Intelligence Commission later formed a subcommittee to develop a draft framework for testing AI in corporate governance.
The draft legislation was described in June 2026 as likely to be introduced in the state legislature the following year. Its status remains a proposal. The committee’s document review and Delaware’s 30-month sandbox plan are the stated next steps in assessing how AI-run companies might operate and where responsibility would fall.
Related Business Coverage
For another example of how business rules affect access to capital, read about bank charters and women entrepreneurs.
For the source reporting, see Ars Technica’s report on Delaware’s AI company plan. For related Grit Daily coverage, see the 777 Partners founder’s guilty plea and a study of startup fraud risks.
