USA Today becomes the latest publisher to sue OpenAI
USA Today Co. and several newspapers it owns are suing OpenAI, alleging the company copied hundreds of thousands of articles to train its AI models without permission. The publisher is seeking more than $250 million in damages, saying the alleged use continues to harm its outlets.
The lawsuit includes The Tennessean, Indy Star, The Columbus Dispatch and The Oklahoman, among others. It joins copyright claims brought by publishers and media organisations including The New York Times, CBC/Radio-Canada and a coalition of nearly 400 local newspapers; OpenAI had not immediately commented.
- USA Today Co. is seeking more than $250 million in damages.
- The lawsuit alleges OpenAI copied hundreds of thousands of articles.
- The case adds to a growing set of copyright actions against OpenAI.
New here? Start with this
OpenAI is a company that has developed artificial intelligence systems, most notably ChatGPT, which can generate human-like written responses to questions and prompts. These AI models are trained using large datasets of text collected from the internet, including articles, books, websites and other published content, from which they learn to predict and generate language patterns. The process of gathering and using this training data has raised questions about copyright and permission.
News publishers rely on readers and subscribers for revenue, so they have a strong financial interest in controlling how their work is used and distributed. Publishers argue that when AI companies train their systems on copyrighted articles without permission or compensation, it violates their intellectual property rights and could potentially harm their business by allowing AI systems to reproduce their insights and content. This concern has prompted multiple major news organisations and coalitions of smaller publishers to take legal action.
The copyright questions raised by AI training have implications beyond journalism, affecting how emerging technologies can be developed and what rights creators hold over their work. The legal outcomes of these cases could establish important precedents for how artificial intelligence companies must operate in relation to copyrighted content.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Newspapers invested substantially in original journalism, and copyright law exists to protect creators' work and ensure they're compensated for it. OpenAI copied hundreds of thousands of articles to train commercial products without permission, directly competing with publishers' own offerings and eroding their revenue. Copyright protections should apply to content used at scale for commercial AI training just as they do in other industries.
The case against
AI training on publicly available data represents transformative use similar to search engines and indexing technologies, which fall under fair use doctrine. The content isn't being republished or directly competing with original articles; rather, it's used to learn patterns and improve AI capabilities that benefit society broadly. Overly restrictive copyright interpretations could significantly impede beneficial technological innovation and progress.