S. 2367 Senate Bill 119th Congress

A bill letting people sue over data used without consent — counting both training on it and generating content derived from it

U.S. Senate Latest update Jul 21, 2025

S.2367, the AI Accountability and Personal Data Protection Act, establishes a federal tort relating to the appropriation, use, collection, processing, sale or other exploitation of individuals data without express, prior consent. Beyond training a generative AI system, it counts generating content that imitates, replicates or is substantially derived from an individual data.

Bill overview (primary data)

  • Bill numberS. 2367
  • TypeSenate Bill
  • Congress119th Congress
  • Latest actionRead twice and referred to the Committee on the Judiciary.(2025-07-21)

Key points

  • S.2367 establishes a federal tort for appropriating, using, collecting, processing, selling or otherwise exploiting individuals data without express, prior consent.
  • It takes the shape of a right for individuals to sue rather than oversight by an agency.
  • Use covers both training a generative AI system and generating covered data pertaining to an individual.
  • The generation branch reaches content that imitates, replicates or is substantially derived from the data.
  • Introduced July 21, 2025 by Mr. Hawley with Mr. Blumenthal and referred to the Judiciary Committee; 94 of the 120 bills this site holds as of 2026-09-02 are at that stage.

1Creating a right to sue rather than a regulation

Legislation on data use comes in two shapes: giving an agency oversight, or giving individuals a right to sue. This bill is the second, establishing a federal tort. It does not depend on an agency enforcement capacity or budget, since the party affected can go to court directly. In exchange, whether to sue rests on the individual judgment and burden.

2What counts as use

Using it for trainingGenerating similar content
Training a generative AI system that the provider sells, rents, licenses or otherwise usesGeneration by a generative AI system of covered data pertaining to an individual
Use on the input sideUse on the output side
The issue arises when data is taken inThe issue arises when content imitating, replicating or substantially derived from it is produced

The core of this bill is including the right-hand column within use. Asking only whether data was used in training runs into the difficulty that training is hard to observe from outside and hard to prove. Capturing the output side — that content imitating, replicating or substantially derived from a person data was generated — lets the affected party start from something they can see.

3The phrase substantially derived

That the text writes imitates, replicates or is substantially derived in three steps signals that something short of an exact copy can fall within scope. It reads as reaching reproduction in a form that is recognizable without being a copy — a voice, a writing style, a manner. Where the line falls will be contested in application, but the intent is set out at the level of the text.

4Sponsors and stage

The sponsors are Mr. Hawley and Mr. Blumenthal, two members of different parties. Introduced July 21, 2025, it was read twice and referred to the Committee on the Judiciary. Of the 120 bills this site holds as of 2026-09-02, 94 (78 percent) remain referred to committee, and this bill is at that stage. At the same date, 70 originate in the House and 50 in the Senate.

Why it matters

A legislative pattern that frames the training-data question as an individual right of action rather than administrative regulation. Reaching content that is substantially derived, not only copied, bears directly on how output controls are designed.

FAQ

What does establishing a tort mean?
Creating a legal basis for an affected individual to sue directly in court, rather than having an agency supervise. It does not depend on enforcement capacity, but suing rests on the individual.
Why cover the generation side?
Whether data was used in training is hard to observe from outside and hard to prove. Capturing the output side lets the affected party start from something they can see.

Sources (primary)

Source: Congress.gov (Library of Congress; U.S. legislative materials, public domain). Links go to the official site.

#Congress bills#Generative AI#Personal data#Tort#Training data
Disclaimer: This site independently summarizes and classifies information based on official data sources. Always verify the latest and accurate information with the official sources. Content on finance, health, legal, and security is information, not advice. This site is not an official website of the U.S. government.