Nielsen's Gracenote sues OpenAI over use of metadata in AI training

By Blake Brittain

March 10 (Reuters) - Nielsen's [RIC:RIC:VALCVN.UL] Gracenote, which creates metadata that identifies movies, TV programs and other media, sued OpenAI in Manhattan federal court on Tuesday, alleging that its work was used without permission to train artificial intelligence.

• Gracenote said in the complaint that OpenAI misused its copyrighted material to train ChatGPT to reproduce its content descriptions and identifiers.

• An OpenAI spokesperson said the company's AI models "empower innovation, and are trained on publicly available data and grounded in fair use."

• Gracenote CEO Jared Grusd said in a statement that OpenAI "chose to use decades of our proprietary work without permission to build and sell its models"

• Gracenote requested an unspecified amount of monetary damages and a court order blocking OpenAI from using its data.

• Gracenote said it employs more than 1,000 editors who "painstakingly source, ingest, aggregate, research, edit, write, curate, and link content" for its database of shows and movies.

• The lawsuit says Gracenote was able to prompt ChatGPT to create copies of its identifiers and descriptions of popular television shows including "Breaking Bad," "Game of Thrones," "The Office" and "Saturday Night Live," indicating that OpenAI used them in its training.

• Gracenote has traditionally licensed its metadata to media distributors. The complaint also says it licenses its material to other AI providers for their training. It alleges that OpenAI's use of Gracenote's content threatens to undercut both markets.