Encyclopedia Britannica Sues OpenAI Over Alleged Misuse of Content for AI Training

Encyclopaedia Britannica and its subsidiary Merriam-Webster have filed a lawsuit against OpenAI in a Manhattan federal court, accusing the company of improperly using their copyrighted materials to train AI systems.

In the complaint filed on March 16, Britannica alleges that OpenAI copied nearly 100,000 articles, including encyclopedia entries and dictionary definitions, to train models behind its chatbot, ChatGPT.

The publisher claims the chatbot can produce “near-verbatim” outputs of its content, potentially diverting traffic away from its platforms and harming its business.

Britannica also argued that AI-generated summaries have “cannibalised” its web audience and accused OpenAI of trademark infringement, alleging the system falsely implies authorization and cites Britannica in inaccurate or fabricated responses.

OpenAI responded by stating its models are trained on publicly available data and operate under fair use principles. The company has consistently defended its practices amid growing legal scrutiny.

The case adds to a series of lawsuits filed by publishers and authors against AI firms over training data usage. Britannica is also pursuing a separate legal battle against Perplexity AI on similar grounds.

Britannica is seeking unspecified damages and a court order to halt the alleged infringement.