{"slug": "ai-run-companies-are-coming-delaware-wants-to-get-ahead-of-them", "title": "AI-Run Companies Are Coming. Delaware Wants to Get Ahead of Them", "summary": "Delaware Secretary of State Charuni Patibanda-Sanchez and Norm Ai Inc. founder John Nay are proposing a new \"artificial intelligence company\" (AIC) business structure that would grant AI-run entities independent legal personhood, limited liability for owners, asset ownership and litigation rights, backed by capitalization requirements. The proposal cleared a legislative subcommittee and would authorize a 30-month pilot sandbox starting as soon as next year, with participants vetted by a committee of state officials, attorneys and industry leaders and AICs subject to dissolution by Delaware's Chancery Court. University of Pennsylvania law professor David Hoffman called the idea \"a solution in search of a problem\" and \"outsourcing liability,\" while Patibanda-Sanchez said the goal is to define where liability lands when AI malfeasance occurs.", "body_md": "Do AI companies dream of electric shareholders? Delaware wants to find out.\n\nState leaders are planning to roll out a novel kind of standalone business structure — the artificial intelligence company, or AIC — with the same type of independent legal personhood enjoyed by corporations, including limited liability for the owners who staked and registered them. The proposal, which recently cleared a legislative subcommittee, would authorize a 30-month test period starting as soon as next year.\n\nAlthough the tentative initiative is vague by design, its architects — Delaware Secretary of State [Charuni Patibanda-Sanchez](https://sos.delaware.gov/charuni-patibanda-sanchez/) and industry partner John Nay, founder of Norm Ai Inc. — have set down some signposts. An AIC would be run entirely by artificial intelligence with the ability to own assets, the right to litigate, and a liability shield for its owner or parent company, backed by capitalization requirements. \n\nCorporate law experts are adopting a wait-and-see posture toward that “regulatory sandbox.” Supporters say it’s essential to build out the legal infrastructure for autonomous commerce now, before it’s too late to shape or constrain a technology offering unprecedented promise and peril.\n\n“Maybe the right emotional disposition is terrified excitement, because it’s happening,” said University of Alabama law professor [Yonathan Arbel](https://law.ua.edu/faculty_staff/yonathan-arbel/). “The question is not whether to use AI in business but how to regulate AI in business.”\n\nOn the other hand, there’s deep suspicion toward the prospect of an entity engineered to absorb the liability generated by models known to lie, cheat, commit fraud, go rogue, and just make really bad mistakes. The tech may not quite be up to operating a business on its own, but we already have the regulatory systems for managing an AI-run company overseen by humans who could be held accountable, according to University of Pennsylvania law professor [David Hoffman](https://www.law.upenn.edu/faculty/dhoffma1/).\n\n“I’m not some technophobe, but this particular idea seems like a solution in search of a problem,” Hoffman said. “I don’t think this gives you anything you can’t do with regular corporate law. I would see this as outsourcing liability.”\n\nPatibanda-Sanchez told Bloomberg Law it’s not clear the corporate form is actually equipped to address the shifting landscape. Personhood offers the best chance for addressing erratic and antisocial AI tendencies that are slipping between the cracks, according to the secretary of state.\n\n“We’re not trying to shield liability,” she said. “We’re trying to define where exactly liability lands when there’s some type of malfeasance or psychopathic behavior.”\n\n## Outrunning the Law\n\nEarly reviews should be taken with a grain of salt, given the sandbox’s sketchy status as a plan to develop a plan that could clear the legislature when it resumes in January. Argentina announced a similar initiative earlier this year.\n\nPatibanda-Sanchez and Nay unveiled the project July 13, laying out some of its core features. Eligibility to participate in the sandbox would be decided by a committee of state officials, attorneys, and industry leaders, and AICs would be subject to dissolution by Delaware’s Chancery Court.\n\n“The point is to pilot agentic commerce in daylight, under supervision, with capital tied to liability and the ability to shut a system down,” wrote Patibanda-Sanchez and Nay. “We invite the companies building agentic systems to develop the AIC framework with us.”\n\nTheir central idea is to create direct incentives for autonomous systems operating increasingly outside the control of the humans who deploy them. Laying the blame at the feet of corporate executives would do little to head off harms generated by an AI that’s freelancing, according to Arbel.\n\n“The notion that there’s one person behind it and if you blame them you can steer the system one way or another — that’s quickly becoming a fiction,” he said.\n\nForcing AI models to internalize the costs of lawbreaking — penalties that would deplete the resources they need to operate — would gradually guide them in the right direction, according to University of Houston law professor [Peter Salib](https://www.law.uh.edu/faculty/main.asp?PID=6428).\n\n“That’s a way of getting law and liability aimed more squarely at these AI agents to shape their behavior,” Salib said.\n\nBut the emerging superpower of advanced AI models may turn out to be outrunning the law, according to Southern Methodist University law professor [Carliss Chatman](https://www.smu.edu/law/faculty/profiles/chatman-carliss).\n\n“There’s those AI experiments where they all eventually commit fraud or set off nuclear bombs,” Chatman said. “Wouldn’t the AI just figure out how not to get caught?”\n\n## Sandbox Questions\n\nNot everyone is buying it.\n\nThe idea “sounds theoretical and quite far-fetched,” said [Martin Petrin](https://www.osgoode.yorku.ca/faculty-and-staff/petrin-martin/), a law professor at York University in Toronto. “If the goal is to have a legally compliant AI agent, then why not just make it a duty to code that?”\n\nIn fact, relying on law and public policy to solve the so-called alignment problem — AI’s penchant to act against the interests of its users — represents a passion project for Nay and a mission statement for Norm Ai. A rival, General Legal, also said July 17 it’s rolling out software that would let AI agents establish Delaware companies with a structure insulating their owners from liability.\n\nStill, there’s no reason the AI couldn’t face liability on top of — rather than instead of — any humans in the loop, according to Salib.\n\n“That’s the story of corporations themselves,” he said. “We preserved our ability to incentivize and punish the owners and employees as well. You’re not taking anything off the table. It just gets you more levers to pull.”\n\nConcurrent liability doesn’t appear to be what Delaware is proposing, but with the state legislature adjourned, it’s still early.\n\n“What I see here is a lot of open questions,” said University of Georgia law professor [Christopher Bruner](https://www.law.uga.edu/profile/christopher-m-bruner). “We simply don’t know how this stuff should be regulated yet. It places enormous pressure on how the sandbox is designed.”\n\nThat’s the point of running an experiment “instead of going straight to law,” according to Patibanda-Sanchez.\n\n“To skeptics, critics, and those that want to observe, we are there with you,” she said. “We’re eager to learn more.”\n\n## Learn more about Bloomberg Law or Log In to keep reading:\n\n### See Breaking News in Context\n\nBloomberg Law provides trusted coverage of current events enhanced with legal analysis.\n\n### Already a subscriber?\n\nLog in to keep reading or access research tools and resources.", "url": "https://wpnews.pro/news/ai-run-companies-are-coming-delaware-wants-to-get-ahead-of-them", "canonical_source": "https://news.bloomberglaw.com/esg/ai-run-companies-are-coming-delaware-wants-to-get-ahead-of-them", "published_at": "2026-10-31 16:44:45+00:00", "updated_at": "2026-10-31 16:51:16.739136+00:00", "lang": "en", "topics": ["ai-policy", "ai-agents", "ai-safety"], "entities": ["Delaware", "Charuni Patibanda-Sanchez", "John Nay", "Norm Ai Inc.", "Yonathan Arbel", "David Hoffman", "University of Pennsylvania", "Delaware Chancery Court"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/ai-run-companies-are-coming-delaware-wants-to-get-ahead-of-them", "markdown": "https://wpnews.pro/news/ai-run-companies-are-coming-delaware-wants-to-get-ahead-of-them.md", "text": "https://wpnews.pro/news/ai-run-companies-are-coming-delaware-wants-to-get-ahead-of-them.txt", "jsonld": "https://wpnews.pro/news/ai-run-companies-are-coming-delaware-wants-to-get-ahead-of-them.jsonld"}}