House Oversight Chair Comer Questions Delaware’s Proposal for AI-Run Companies

Table of Contents

• You might want to know
• Main Topic
• Key Insights Table
• Afterwards...

You might want to know

• Could Delaware’s proposed AI Company structure make autonomous businesses more accountable, or could it shield people behind them from responsibility?
• What information is House Oversight Committee Chairman James Comer seeking about the proposal and its development?

Main Topic

House Oversight Committee Chairman James Comer (R-KY) has opened an investigation into Delaware’s proposal to establish a new kind of company that could be run day to day by an artificial intelligence agent. The inquiry focuses on the proposed Artificial Intelligence Company, or AIC, and asks whether its legal structure could clarify accountability for autonomous systems—or protect the people and investors behind them from liability when those systems cause harm.

In a letter to Delaware Gov. Matt Meyer, Comer said the proposal raises concerns about responsibility, transparency, and the influence of private-sector contributors. He warned that the framework could be used to shield powerful AI companies and their investors from responsibility for reckless harmful actions. He also raised the possibility that opaque interests could seek anticompetitive regulatory capture through state AI initiatives. These are concerns raised by the committee; they are not established findings about the proposal or its contributors.

Comer asked Meyer to provide documents and communications concerning the AIC’s development since January 2025. The requested materials include “internal AI chat logs,” agreements with outside contributors, records of safety concerns, and all drafts of the proposal. The deadline stated in the letter is November 9. The committee said the information could help inform federal legislation, including proposals that would limit state AI regulatory initiatives.

The request follows questions about how the proposal was developed and what safeguards informed it. The committee’s letter refers to a December 2025 report by Delaware’s AI Commission, which, according to Comer, did not mention efforts to assess safety risks or consult independent experts. It also points to the Delaware Secretary of State’s office declining to tell reporters about its working relationship with an AI company involved in the project. These points form part of the committee’s rationale for seeking records about the proposal’s origins and oversight.

The AIC proposal was presented publicly in a Fortune op-ed in July by Delaware Secretary of State Charuni Patibanda-Sanchez and John Nay, CEO of legal AI firm Norm Ai. They described it as a public-private partnership led by Norm Ai. Under the proposed structure, an AIC could sue and be sued, own property, and take on obligations at the direction of its AI agent. In other words, the entity would have a legal identity capable of participating in ordinary legal and commercial relationships, even as an AI system carried out its operations.

The plan would give each AIC a single member responsible for ensuring that it remained adequately capitalized. That member would generally be shielded from the company’s debts, unless the member failed to capitalize the entity or used it to commit fraud or a willful violation of law. This arrangement is central to the debate: supporters describe a defined legal boundary for an AI-operated business, while critics worry that protections for the member could weaken accountability if an AI agent causes damage.

The proposed entities would be limited to a 30-month regulatory sandbox overseen by a committee of state officials and outside experts. AICs would have to disclose their status to counterparties and would be barred from banking. The sandbox is intended to provide a supervised setting in which the model could be tested, though the quality of supervision, the range of activities permitted, and the handling of harms would be important questions for lawmakers and regulators.

The proposal’s authors said their goal was to “wrap AI in legal form.” They argued that giving an autonomous system a recognizable legal identity could make its conduct “visible, traceable, and accountable.” The reasoning is that a formal entity could create a clearer point of reference for contracts, obligations, disputes, and oversight. However, establishing a legal identity does not, by itself, resolve questions about who should answer for an AI agent’s decisions or how victims could obtain remedies.

Some legal scholars have questioned whether the proposal provides enough of a broader regulatory framework. Christopher Bruner, a business law professor at the University of Georgia, told WHYY in September, “There’s really no overarching regulatory plan here for dealing with the fallout that AI-governed companies could bring.” His comment highlights a distinction between creating a legal vehicle for an AI-run organization and addressing the wider risks that could arise from its conduct. Legal status may help organize responsibility, but it may not replace rules governing safety, transparency, or specific business activities.

Patibanda-Sanchez has disputed the characterization of the proposal as a liability shield. Speaking to Bloomberg Law in July, she said, “We’re not trying to shield liability. We’re trying to define where exactly liability lands when there’s some type of malfeasance or psychopathic behavior.” Her explanation frames the AIC as a way to identify where responsibility belongs, rather than eliminate it. Whether the proposed design would achieve that outcome will depend on its final terms and how those terms are enforced.

Meyer has also defended the effort. On a WHYY call-in show in September, he said, “It is coming, and the solution is not ignoring it.” State lawmakers are expected to consider draft legislation when they return in January. That next stage could determine whether the proposal advances, what provisions are revised, and how officials respond to concerns about liability, safety, independent expertise, and the influence of outside contributors.

The inquiry therefore places the proposal within a wider debate over how governments should respond to AI systems that can make decisions or carry out work with limited day-to-day human involvement. A legal structure could make an AI-run company easier to identify and subject to certain obligations. At the same time, any limits on the liability of the person or entity behind it could raise questions about whether the structure provides meaningful recourse when harm occurs. Those questions will require careful attention to the distinction between the AI agent, the AIC itself, its member, and any organizations involved in developing or supervising the system.

For now, the committee’s letter is a request for information, not a conclusion that Delaware’s proposal is unlawful or that its contributors acted improperly. The materials Comer requested may help clarify how the idea was drafted, what safety issues were considered, and how public officials and private-sector participants worked together. Delaware’s lawmakers will then face the task of weighing the proposed benefits of a supervised legal framework against the risks identified by critics and federal investigators.

Key Insights Table

Afterwards...

The next steps will depend on what the requested records reveal and how Delaware lawmakers shape the draft legislation. A durable framework will need to explain not only how an AI-run company acquires legal status, but also how responsibility is assigned when its actions cause harm, what safeguards apply, and how independent oversight works in practice. The committee’s inquiry and the state’s legislative process may bring those questions into sharper focus. The central challenge is to support clear, enforceable accountability without assuming that a corporate form alone can manage every risk associated with autonomous systems.