Legal Scholars Propose Ending Limited Liability for AI Firms

Legal scholars from Columbia Law School and Yale Law School are proposing the removal of limited liability protections for artificial intelligence companies facing mass torts. The proposal follows warnings from Dario Amodei, CEO of Anthropic, as well as Sam Altman and Elon Musk, that AI agents could cause hundreds of billions of dollars in damage within a single year.

The scholars argue that shareholders and controlling owners, including Amodei and Mark Zuckerberg, should be held personally liable for damages if their companies go bankrupt. This shift would create a direct financial incentive for AI laboratories to prioritize and solve safety engineering problems.

To implement these standards, the authors suggest the U.S. government could enact new rules or make American chip access conditional on Chinese AI firms adopting similar liability frameworks. As a potential industry precedent, the proposal suggests that Anthropic could waive limited liability for torts before the company goes public.