The chairman of the US House Committee on Oversight and Government Reform has opened an investigation into Delaware's proposal to create a new kind of company run by AI agents, warning that the structure could let human owners dodge legal responsibility when the software causes harm.
In a letter to Delaware Governor Matt Meyer, US Rep. James Comer (R-Ky.) said the committee "seeks assistance obtaining documents related to your state administration's efforts to establish a new form of corporate entity to be known as an Artificial Intelligence Company (AIC)," as Ars Technica reported. Comer chairs the Oversight committee and said the proposal is "ripe for abuse."
What Delaware is proposing
The AIC plan, unveiled in July 2026 by Delaware Secretary of State Charuni Patibanda-Sanchez, would establish a corporate entity whose day-to-day affairs are managed by an AI agent rather than a person. A draft bill that may be submitted to the Delaware State Legislature says an AIC "may have any lawful purpose, regardless of whether for profit, except for the business of banking."
Under the draft, AI agents could be "designated to manage the business and affairs, and to conduct the business, of the AIC." The legislation also proposes a 30-month "regulatory sandbox" to test the concept. Allowing AI-run companies, supporters argue, would bring more tax and fee revenue to Delaware, which is already the top US destination for business incorporation.
"Wrap AI in legal form"
Writing in Fortune in July 2026, Patibanda-Sanchez and her co-author argued that "nothing in the law of several states clearly prevents an AI agent from operating a company with no human at the helm," and that states should build legal frameworks for autonomous software that can contract, pay suppliers, and make transactions with no human involvement.
"The answer is not to try to prohibit the activity; any effort to do so is just as unlikely to succeed as the technology and its enablers are likely to push the boundaries of the law," the op-ed said. "The answer is to wrap AI in legal form. Give an autonomous system a recognizable legal identity and you make it legible to law. You create a defined target to which responsibility and damages can attach."
The mechanics matter for the liability debate. A Delaware AIC would be a separate legal entity that can sue and be sued in its own name, hold and dispose of property, and incur obligations at the direction of its AI manager. It would have a single member — a person or entity — responsible for keeping the AIC adequately capitalized, and the AIC must keep a log of its activities. That member is shielded from the AIC's debts.
Comer's warning: liability evasion and regulatory capture
Comer's letter argues the structure could work in the opposite direction of the op-ed's accountability rhetoric. "We are concerned that Delaware's AIC proposal could be used to protect powerful AI companies and their investors from responsibility for reckless harmful actions, while also protecting non-transparent forces seeking anticompetitive regulatory capture behind state AI initiatives," the letter said.
Citing the committee's jurisdiction over intergovernmental relationships, Comer requested documents on how the proposal was developed and its potential AI safety risks. He also signaled that Congress may act: the probe would "inform possible federal legislative efforts including proposals to appropriately limit state AI regulatory initiatives" — language that raises the prospect of federal preemption of state AI laws.
A test case for autonomous business
The fight is a preview of the questions that arrive once software, not people, signs contracts and runs operations. If an autonomous company injures someone, breaches a contract, or launders money, plaintiffs currently reach the humans behind it; a structure that formalizes "no human at the helm" could complicate that at scale, especially in Delaware, where a majority of US public companies and Fortune 500 firms choose to incorporate.
The 30-month sandbox design suggests Delaware wants to observe before committing: companies admitted to the program would operate AI-managed under supervision, generating a record of how agents actually behave in corporate roles — what they sign, what they spend, and what goes wrong.
The federalism fault line
Comer's hint at federal legislation to "appropriately limit state AI regulatory initiatives" adds a constitutional layer. A state-chartered corporate form built around AI agents would become a test of how far Washington will go to override state experiments — notable coming from the party that has generally argued against heavy-handed AI rules at the federal level.
The Oversight committee has not set a deadline for Delaware to produce the documents, and the draft bill's fate in the state legislature is unclear. But with Congress now requesting paperwork, the sandbox experiment is already being litigated in the court of federal oversight — before a single AI-managed company exists.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
