OpenAI, Inc., Copyright Infringement Litigation
The causation theory in MDL-3143 is based on the legal argument that training large language models on copyrighted datasets involves copying protected works, which constitutes direct infringement. The process of ingesting and transforming copyrighted data into models like GPT-3 and GPT-4 is alleged to infringe reproduction rights, causing harm by depriving plaintiffs of control over their works and enabling ongoing infringement through generated outputs. Courts recognize that substantial copying during training can establish causation, linking the act of training directly to the infringement and resulting harm.
18
Pending actions
19
Total actions filed
Active
Status
04/03/2025
Established
Filing deadline
The claims are timely as the case was filed within the three-year statute of limitations based on the discovery rule, with the last filing in March 2026.
Who qualifies
Ownership of copyrighted works that were used by OpenAI in training its models, and that the use was unauthorized, constituting infringement. Plaintiffs must demonstrate that their works were used in training or output generation by OpenAI's models, which involves showing that the works were exposed to the AI training process without authorization.
Products involved
- GPT-3
- GPT-4
Alleged injuries
- Unauthorized use of copyrighted works in training datasets
Bellwether trials
No bellwether trials have been scheduled or reported as of the latest court filings; the case remains in active procedural stages with ongoing rulings and transfer to the Northern District of California.
Settlement landscape
There are no publicly reported settlement estimates or case values for MDL-3143. The case is actively litigated with procedural rulings, and no resolution or settlement figures have been disclosed.
This page is generated from the official JPML pending-MDL report and public court records, refreshed monthly. It is provided for attorney reference and is not legal advice.
