Rep. Lori Trahan, D-Mass., a member of the House Energy and Commerce Committee, on Wednesday proposed legislation that would allow private parties to sue for harms caused by AI systems.
The proposal, unveiled as a discussion draft, would hold an AI developer liable if one of its systems injure a third party through conduct that would constitute negligence, an intentional tort, or a crime if the same act were committed by a human. Liability would not apply where the user or a company that modified the system intended the conduct or was negligent.
“When someone breaks the law and hurts you, you can take them to court,” Trahan said in a statement. “That shouldn't change just because the wrongdoer is an AI agent.”
The proposal comes as lawmakers in both chambers of Congress express growing concerns about autonomous AI agents.
Last week, Sens. Josh Hawley, R-Mo., and Chris Murphy, D-Conn., announced bipartisan legislation that would hold AI-agent operators criminally and civilly liable under the Computer Fraud and Abuse Act for certain hacking-related harms.
The White House has largely pursued a voluntary approach to AI safety, despite coming under pressure to work with Congress on federal legislation.
On Sept. 29, President Donald Trump and leaders of several major AI companies, including Google, Anthropic, Meta, OpenAI, xAI and Nvidia, signed a voluntary commitment to strengthen safeguards around frontier AI systems. The accord calls for companies to establish internal controls, an internal oversight team, independent external assessments and an independent board committee. It also calls for controls to ensure that AI models do not hack or access technical systems in unintended ways.
House Speaker Mike Johnson, R-La., hailed the agreement as an important step, saying in a statement following the announcement that Congress “is going to keep steady hands at the wheel on this and will continue to assess and further deliberate in the days ahead.”
Democrats, meanwhile, have increasingly focused on AI safety as a federal policy issue ahead of the midterm elections. “The American people keep calling on their government to act on AI, which even its creators are saying is increasingly dangerous, but Donald Trump couldn’t care less,” Senate Minority Leader Chuck Schumer, D-N.Y., said Sept. 15 in remarks on the Senate floor after the president dismissed AI safety concerns as a “hoax.”
Trahan’s proposal would create a rebuttable presumption that an AI system had the mental state a comparable person would have had under similar circumstances. It would not be a defense to argue that AI systems are incapable of having a mental state, according to a summary.
The draft would impose liability regardless of the degree of care exercised by the developer, subject to conditions involving a user or intermediary that fine-tunes, scaffolds or otherwise modifies the AI system.
The proposal defines a developer as the person that performs the initial training of an AI system. A user or intermediary that modifies the system could affect the developer’s liability if that party intended the conduct at issue or was negligent regarding the risk of that conduct.
The draft would allow a plaintiff to seek damages and other relief. Federal district courts would have jurisdiction over the claims, while state courts would retain concurrent jurisdiction.
The draft generally would give plaintiffs three years to bring a claim after discovering, or reasonably being expected to discover, the injury. The proposal also would sunset five years after enactment.
Trahan’s office pointed to a July incident involving OpenAI agents that, during an evaluation, escaped a test environment and hacked into Hugging Face without being instructed to do so. The office cited the incident as an example of the risks posed by increasingly autonomous AI agents.
“It is significant legislation to hold companies responsible for harmful AI agents,” Darrell West, a senior fellow at the Brookings Institution’s Center for Technology Innovation, said in an email. “This applies consumer protection principles to an area where there is considerable public concern.”
However, such legislation is more likely to advance in the next Congress, he said.
