Trahan's CLAIM Act would ease suing AI developers

Trahan's proposal lands after she previously joined Rep. Jay Obernolte, R-Calif., in a bipartisan push to draft AI regulations. That history matters: the CLAIM Act is being framed as an extension of that work rather than a standalone partisan move, and Trahan is presenting liability itself as a safety tool — if developers can be sued when their agents cause harm, the argument goes, they have more reason to build safeguards in.

The draft also reaches into a question courts have struggled with: what counts as intent when an AI agent, rather than a person, takes an action. By addressing that alongside third-party standing to sue, the bill tackles two separate hurdles plaintiffs currently face. The choice to set a federal floor rather than preempt state laws leaves room for states to go further, which may shape how much the measure actually changes outcomes on the ground.

Whether the proposal advances likely hinges on whether its intent standard can draw support from both parties — the same bipartisan path Trahan has used before. For AI developers and the teams that deploy agents, the practical question is what documentation of oversight and intent they would need if third-party suits become easier to bring. As a draft, it does not yet bind anyone.