This US state wants to let AI set up and run an entire business — with only a single human needed

The state of Delaware wishes to add businesses run by AI agents to the already lengthy list of generous regulations in operation Rep.

<![CDATA[ <article> <ul><li><strong>The state of Delaware wishes to add businesses run by AI agents to the already lengthy list of generous regulations in operation</strong></li><li><strong>Rep. James Comer is demanding safety assessments and warning of potential federal legislation if the results are not satisfactory</strong></li><li><strong>Delaware’s bill specifies a 30-month regulatory sandbox for testing, and allows any lawful business to be operated by an Artificial Intelligence Company except banking</strong></li></ul><p>The US state of Delaware is planning to add to its generous business rules by allowing companies to be run by AI agents – but Congress lawmakers are not so sure, with one Republican representative writing to the state's governor expressing his concerns, with demands for safety assessments.</p><p>Businesses with a Delaware address operate under a range of advantageous regulations, including low (or zero, depending on where the businesses conducts its sales) tax rates and specialized courts.</p><p>The prospect of an Artificial Intelligence Company (AIC) being founded is not unlikely, but law experts believe that Delaware’s bill could be used to implement a liability shield for businesses with AI interests.</p><h2 id="harmful-actions">Harmful actions</h2><p>Delaaware’s plans have come under the scrutiny of the House Oversight Committee, wish its chairman, Rep. James Comer, pushing back, with concerns over the ability to place an AI in charge of a business. </p><p>According to the <a href="https://www.ft.com/content/d5021cdf-816e-4cb9-bbf9-e4ba816a004f" target="_blank">Financial Times</a>, Comer wrote to the Delaware governor Matt Meyer.“We are concerned," he wrote, "that Delaware’s AIC proposal could be used to protect powerful AI companies and their investors from responsibility for reckless harmful actions.” </p><p>In short, if things go wrong, the AI is to blame and the owner of the business has no responsibility.</p><p>The draft bill states a 30-month regulatory period in which the AI would be sandboxed. During this time, the concept can then be tested, with an AI agent making the business decisions rather than a human. Like any other business, the company could be sued, and take legal action itself if warranted.</p><p>Delaware’s plan could aid in the establishing of AI-centric corporate structures (such as autonomous commerce businesses), something that many believe could happen overseas if no US state acts quickly.</p><h2 id="legal-certainty">Legal certainty</h2><p>Various names in business and AI have previously supported the idea of AI-run businesses. Most notable is Mark Cuban, the billionaire investor who <a href="https://x.com/mcuban/status/2048813656108863893" target="_blank">suggested</a> earlier this year that as “First to market makes a killing,” any “governor of a state looking to grow their revenues” should allow “anyone to incorporate using an AI agent and stablecoins.”</p><p>Andrew Baker, a professor of law at the University of California, indicated to the <em>FT</em> that the Delaware legislation could be an attempt to reach a midway point, perhaps to achieve “legal certainty” for Delaware-registered businesses using agentic AI.</p><p>“Ultimately the most likely outcome appears to be an overly generous liability shield which seems unwarranted without some underlying policy rationale that hasn’t been developed or justified well,” Baker said. </p><p>As corporations operate as a shield to personal liability, this would not seem beyond the reach of business legislation. However, those cases involve businesses run by humans who have to be accountable to themselves if not the law, and an AI agent has no such moral restrictions.</p> </article> ]]>