USA Today Company, the American newspaper giant formerly known as Gannett, has filed a $250 million lawsuit against ChatGPT developer OpenAI over allegations of copyright infringement.
The lawsuit filed this week claims OpenAI used copyrighted material from USA Today and its network of local newspapers. More than a dozen USA Today-owned newspapers are listed as plaintiffs in the case, including the Arizona Republic, the Milwaukee Journal Sentinel, the Des Moines Register and the Tennessean.
The publishers claim OpenAI used its protected materials to train ChatGPT, including news articles that were supposed to be protected by paywalls. According to a compliant filed in federal court on Thursday, OpenAI’s WebText training dataset contains more than 160,000 entries originating from websites operated by USA Today-owned publications, including nearly 83,300 newspaper articles from USA Today’s flagship website.
USA Today claims ChatGPT harms its business because it produces summaries of news reports that are pulled from its websites and other digital platforms. In doing so, the newspaper giant says ChatGPT is directly responsible for a significant drop in web traffic and advertising revenue.
The complaint filed by USA Today’s attorneys cite comments made by Nick Turley, a computer scientist at OpenAI who is in charge of ChatGPT, who allegedly said that OpenAI’s artificial intelligence-generated, summarized answers could cause a drop in web traffic to other websites.
The complaint also references an OpenAI engineer who wrote that users would be unlikely to follow source links regardless of how prominently those links appeared.
OpenAI has not commented on the lawsuit.
USA Today is the latest major news outlet to sue an artificial intelligence developer over their purported use of copyrighted information to train their large language models (LLMs), the underlying technology that makes conversational chatbots like ChatGPT and other artificial intelligence tools work.
The New York Times has open lawsuits filed against Microsoft, OpenAI and Perplexity, claiming their news articles were used to train LLMs without permission. At the time the lawsuit was initially filed in 2023, Microsoft had made financial investments in OpenAI.
Dow Jones, the parent company of the Wall Street Journal and the New York Post, also have an open lawsuit against Perplexity. Penske Media, the parent company of Variety and Deadline Hollywood, unsuccessfully sued Google over its use of artificial intelligence-generated summaries in search results.
Some of those publishers have banded together to push for tougher laws that would restrict some use of news content to train artificial intelligence models. One law under consideration would require tech companies honor a request by news publishers to block the viewing of their content by artificial intelligence robots, and would allow federal investigators to bring civil actions for violations of certain proposed rules. That federal legislation is modeled off a similar law in New York state, which lawmakers approved earlier this year. New York Governor Kathy Hochul has yet to sign the bill.
More Stories
- Comscore: ChatGPT losing market share to Gemini, Claude
- Google pays some news publishers as AI-generated results dominate search
- New York lawmakers approve bill limiting AI crawling of news articles
- Variety owner Penske loses lawsuit against Google over lost search traffic
- Publishers push for new law banning bot-based AI scraping
