Delaware is considering a legal structure for companies whose business decisions are handled by AI agents, and a senior US lawmaker is asking questions about the plan. The proposal could give AI-run firms legal standing while generally shielding their sole member from company liabilities, making accountability the central issue rather than a sci-fi flourish.
Watch Desk analysis
What happened
TechSpot reports that House Oversight Committee chairman James Comer wrote to Delaware Governor Matt Meyer on 26 October requesting documents about the proposed Artificial Intelligence Company, or AIC. Comer asked about safety assessments and industry involvement, and warned the inquiry could lead to federal legislation restricting state AI initiatives.
Under the draft bill, Delaware would create a 30-month regulatory sandbox. An AIC could pursue lawful purposes other than banking, with an AI agent making business decisions without routine human supervision. The company could sue and be sued, and would have one member responsible for keeping it adequately funded. That member would generally be protected from the company’s liabilities, with exceptions including fraud and knowingly allowing activity outside the company’s stated purpose. TechSpot says the proposal remains a plan, with a bill likely to be introduced next year.
Why it matters
This is a live question about who answers when software makes consequential decisions inside a company. Supporters argue that a separate legal identity could make an AI-run firm’s actions traceable and provide a target for liability. Critics quoted by TechSpot question whether the proposed protections for the member go too far.
That tension has practical consequences: a regulatory sandbox could let the model be tested, but the proposed liability rules would shape who bears the cost if things go wrong. Delaware also has a financial incentive to attract company formations, TechSpot notes, though that alone does not establish why the proposal was made.
Our read
The proposal deserves attention, but it is not law, and an AI agent running a company is not yet a routine legal reality. The useful debate is more specific than “robots as bosses”: what oversight should be required, and should the person or organisation behind an agent retain responsibility when it acts?
What to watch
- Whether Delaware introduces the bill next year, and what changes before then.
- What Meyer’s office says in response to Comer’s letter.
- Whether the proposed liability exceptions satisfy critics concerned about accountability.
Discussion spark: If an AI agent makes a harmful business decision, should the company’s member remain liable even when the draft law’s exceptions do not apply?
Sources and evidence
- Source update (28 October 2026, 12:49 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.