OpenAI is accused of plagiarism: British encyclopedia initiates legal proceedings after ChatGPT nearly fully reproduced its materials

Court battle: “Britannica” and Merriam‑Webster file lawsuit against OpenAI

*Brief overview of the case*

- Plaintiffs – “Britannica” (Encyclopedia Britannica) and publisher Merriam‑Webster.

- Claim – compensation for unlawful use of their copyrighted content in AI training, plus cessation of publishing answers that “substantially reproduce” original texts.

*Specific allegations*

1. Unauthorized copying

The lawsuit alleges that OpenAI repeatedly used encyclopedia materials without permission. GPT‑4 “memorized” significant fragments and now produces near‑verbatim copies upon request.

2. Examples of matches

The legal document contains parallels between OpenAI model responses and encyclopedia texts, where entire passages match word for word.

3. Traffic diversion

Britannica accuses the company of “pulling” users: AI answers replace direct access to its website, thereby competing with a traditional search engine.

*Context and prior cases*

- This is not the first lawsuit against OpenAI. A group of publishers has filed several legal actions in recent years.

- Previously The New York Times also initiated litigation, accusing the company of mass copying its articles.

- In September Anthropic settled a class‑action copyright infringement claim over books, paying authors $1.5 billion.

*Conclusion*

The lawsuits illustrate growing tension between AI developers and rights holders. The question remains open: can OpenAI prove the legality of using large volumes of text for training its models, or will the case end in another settlement?