US Rep. James Comer (R-Ky.) opposes a Delaware proposal allowing AI agents to run entire businesses, arguing it would let human owners evade liability for reckless AI actions. Comer, chairman of the House Oversight Committee, warned the plan could shield powerful AI firms from responsibility and enable regulatory capture.
He requested documents from Governor Matt Meyer and suggested possible federal legislation to limit state AI authority. The draft bill would create an Artificial Intelligence Company (AIC) structure, permitting AI agents to manage operations, sue, hold property, and incur obligations. A 30-month regulatory sandbox would test the concept.
Proponents argue wrapping AI in legal form makes it legible to law, creating traceable targets for responsibility. However, critics counter that the "member" shield protects those who fail to capitalize adequately or use the entity for fraud. The proposal excludes banking but allows AI across other industries.
If implemented, the plan could significantly alter corporate governance and liability structures for autonomous enterprises.
Source: Ars Technica · Summarized by HeadlinesBriefing