cryptojournalist
Judge Dismisses the Case, but It Could Return
A federal judge in New York recently dismissed a lawsuit against OpenAI, the company behind ChatGPT, over its use of news articles to train its AI. Filed by news sites Raw Story and AlterNet, the lawsuit claimed that OpenAI trained ChatGPT on their articles without permission or payment. The judge, Colleen McMahon, decided the plaintiffs hadn’t shown clear financial harm. But she left the door open for them to refile if they can bring more evidence to support their claims.
Key Issue: Fair Use or Fair Pay?
At the heart of the case is a big question: Should OpenAI be allowed to use copyrighted content for free when training ChatGPT? Judge McMahon pointed out that the lawsuit isn’t just about keeping articles out of OpenAI’s training data, but more about the fact that the news sites weren’t paid. Right now, the judge feels the plaintiffs haven’t made a strong enough case, but that could change if they refile with more solid evidence.
Not OpenAI’s First Copyright Challenge
This isn’t OpenAI’s only copyright battle. Major publishers like The New York Times, Time, and The Associated Press have all taken legal action, claiming that OpenAI used their articles to train ChatGPT without permission. These cases are part of a broader debate about how AI companies should handle copyrighted materials and if they should pay for access.
Building Partnerships with News Outlets
Amid these legal battles, OpenAI has started forming official partnerships with news outlets to legally access their content. It now has agreements with the Financial Times, France’s Le Monde, Spain’s Prisa Media, and Germany’s Axel Springer. This way, OpenAI can offer more accurate news through its ChatGPT Search feature, launched on November 1, which answers user questions with up-to-date news.
Meta Also Jumps In
OpenAI isn’t alone in trying to work with news organizations. On October 25, Meta—the parent company of Facebook and Instagram—announced a deal with Reuters. This allows its own AI to link directly to Reuters articles in response to user questions about current events.
