Delaware lawmakers are weighing legislation that would allow autonomous AI agents to manage entire corporate entities, drawing immediate pushback from federal officials over corporate liability. The draft proposal introduces a 30-month regulatory sandbox permitting software models to run commercial operations with minimal human involvement across almost every sector.
Key points
- Delaware has drafted legislation creating an Artificial Intelligence Company status that allows software models to make corporate management decisions.
- The plan includes a 30-month testing sandbox and permits any lawful business except commercial banking.
- US House Oversight Committee Chairman James Comer has warned Delaware Governor Matt Meyer that Congress may intervene if safety assessments fail.
- Legal scholars argue the framework risks shielding corporate owners from accountability when automated systems cause harm.
The mechanics of the AI company proposal
Delaware serves as the legal home for more than half of publicly traded companies in the United States, largely due to its specialised Court of Chancery and predictable corporate code. Under the proposed draft measure reported by the Financial Times, the state would extend those corporate structures to autonomous entities designated as an Artificial Intelligence Company, or AIC.
The bill requires only a single human to establish the entity before handing executive control to autonomous AI agents. The framework then initiates a 30-month regulatory sandbox where the automated entity can conduct regular business operations. Under the proposed rules, the AI entity possesses legal personhood: it can enter contracts, sue counter-parties, and face lawsuits in court.
Delaware drafters excluded traditional banking from the scope of the bill, but left virtually every other legal commercial enterprise open to algorithmic management. Proponents argue that the measure will prevent corporate innovation from fleeing to offshore jurisdictions that are preparing similar autonomous corporate rules.
Congress raises liability warnings
The prospect of algorithmic directors has alarmed federal lawmakers. Representative James Comer, chairman of the House Committee on Oversight and Accountability, sent a formal letter to Delaware Governor Matt Meyer challenging the initiative and demanding safety assessments.
Comer warned that federal legislation could preempt Delaware’s state statute if the framework compromises consumer protection. In his letter, Comer stated that the committee is concerned “that Delaware’s AIC proposal could be used to protect powerful AI companies and their investors from responsibility for reckless harmful actions.”
The central point of friction is accountability. While traditional corporate charters protect investors through limited liability, human directors still face fiduciary duties and regulatory enforcement. Placing software at the head of a corporate entity introduces legal ambiguity when algorithms commit fraud, breach contracts, or cause financial damage, leaving injured parties with no human target for legal restitution.
Debate over corporate certainty and investor shields
The push for algorithmic incorporation has attracted high-profile backing in the technology sector. Investor Mark Cuban publicly argued earlier this year that states could generate substantial corporate tax revenues by allowing founders to register companies using software agents and stablecoins, stating that “first to market makes a killing.”
The corporate legal sector remains divided on whether statutory changes are necessary. The emergence of enterprise AI agents in private commerce has already automated complex corporate tasks, while cloud platforms expanding agent automation handle administrative operations at scale. However, formal governance by algorithms represents an untested legal boundary.
Andrew Baker, a professor of law at the University of California, told the Financial Times that Delaware appears to be seeking regulatory certainty for domestic businesses. Yet Baker cautioned that the current approach risks establishing an unjustified legal shield that shelters company owners behind automated decision-makers.
What happens next
Delaware state officials have not set a formal date to introduce the AIC bill to the General Assembly. Governor Meyer’s office must now decide whether to amend the draft language to address the oversight concerns raised by Comer or proceed with the testing sandbox as written.
