Delaware’s proposal to create a new corporate form run by artificial intelligence is drawing scrutiny from legal experts who say the state has not adequately answered who is accountable when an autonomous company causes harm.
The proposal, unveiled over the summer by Delaware Secretary of State Charuni Patibanda-Sanchez, would create an “artificial intelligence company,” or AIC, in which an AI agent could manage business operations without human supervision, PYMNTS reported July 29. An AIC could sign contracts, own property, incur obligations, and sue or be sued in its own name. The banking business would be excluded.
Supporters view the structure as a way to bring autonomous AI agents inside the traditional legal system rather than allowing agentic commerce to develop outside established mechanisms for liability and judicial oversight, the report said. The proposal would initially operate through a 30-month regulatory sandbox overseen by Delaware officials and outside experts.
However, legal scholars said giving an AI system a corporate identity does not necessarily solve the accountability problem and could instead eliminate some of the mechanisms corporate law traditionally uses to control risky behavior, Route Fifty reported Thursday (Sept. 24).
“The main issue is really accountability because, in regular corporations, we have humans,” York University law professor Martin Petrin said, per the report.
Directors, officers, managers and employees can be constrained by fiduciary duties, face personal liability in some circumstances and suffer reputational consequences for misconduct, he said in the report. An autonomous AI has none of those incentives.
Petrin said Delaware appears to be “jumping ahead straight to the future” by moving beyond proposals to put AI directors on boards or use AI alongside human management and instead contemplating corporations “with no human supervision,” per the report.
The draft does contain safeguards. A person or organization sponsoring an AIC would have to ensure it is “adequately capitalized” to cover expenses, liabilities and debts, the report said. AICs would also have to disclose their autonomous nature and maintain logs of their activities.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
Those provisions, however, leave questions unanswered, said Hofstra University associate law professor Sergio Alberto Gramitto Ricci, according to the report. The proposal does not specify how much capital would be adequate, for example, or provide a detailed governance framework for relationships between an AI-controlled company and human employees or managers.
“How would managers report to the AI?” Ricci asked in the report. “These are aspects of governance that really make a corporation work.”
The objections go to the heart of the proposal’s rationale. Patibanda-Sanchez and Norm Ai CEO John Nay said putting autonomous agents inside a recognizable corporate structure would create an identifiable entity against which responsibility and damages could attach, PYMNTS reported in July. The AIC could therefore make autonomous activity more visible and traceable while giving courts an entity over which they could exercise jurisdiction.
Critics are effectively asking whether an entity capable of paying damages is enough if there is no human decision-maker subject to fiduciary duties and other behavioral constraints.
The stakes are unusually high for Delaware because corporate law is both a major source of revenue and an important part of the state’s economic identity, the Route Fifty report said. More than 2 million companies are incorporated there, generating roughly $2 billion in taxes and fees. Additionally, more than two-thirds of Fortune 500 companies are incorporated in the state.
Patibanda-Sanchez has said a new AI corporate form could provide “a huge benefit to our state’s revenues,” according to the report. Ricci warned, however, that failures or harms involving autonomous companies could undermine Delaware’s reputation for predictable corporate governance.
“If things go well, it might be a marginal competitive advantage for Delaware,” Ricci said, per the report. But if AICs cause significant harm, Delaware “might lose its gravitas as the state that moves in a conscious way.”
The proposal is expected to come before lawmakers after the Delaware legislature reconvenes in January, the report said. Its experiment could have implications beyond corporate law. Petrin said other governments are likely to watch closely as they confront similar questions about autonomous and algorithmic decision-making in their own operations.
For all PYMNTS AI coverage, subscribe to the daily AI newsletter.
Add as Preferred Source
