HR 7209: TRAIN Act

HR 7209 in plain English: This bill is early in the legislative process and detailed text is not yet available. Sponsor: Rep. Dean, Madeleine [D-PA-4] (D) · Status: Referred to the House Committee on the Judiciary.

Stated purpose

The TRAIN Act creates a legal process allowing copyright owners to subpoena AI developers to find out whether their copyrighted works were used to train generative AI models.

Arguments supporters make

• Creators currently have no clear way to find out if their work was used to train AI without their permission, and this bill gives them a basic tool to get that information.
• Transparency about what goes into AI training data is a fair first step that lets copyright owners decide whether to pursue further legal action based on actual facts.
• The subpoena is limited to a requester's own works, so it is narrowly targeted and does not let anyone fish through another person's private data.

Arguments opponents make

• AI companies may hold training datasets containing billions of files, and complying with these subpoenas could be technically difficult, costly, and disruptive to business operations.
• A 'subjective good faith belief' is a low bar for triggering a subpoena, which could open the door to a flood of requests that burden courts and developers even when no infringement occurred.
• This process addresses only information-gathering and does not resolve the underlying legal question of whether using copyrighted works to train AI actually constitutes infringement, leaving the core dispute unsettled.

Tradeoffs

The bill gives copyright owners more information and leverage, but imposes new disclosure obligations on AI developers who may argue that revealing training data raises their own competitive and legal risks. Easier access to information for one side means added legal and operational costs for the other.

Current status in Congress: In committee.