The New York Times has revealed that it’s suing OpenAI and Microsoft as part of a new lawsuit that alleges the companies have been using “millions” of its articles to train their chatbots, something that the publication isn’t too happy about for reasons that include the trending technology’s potential for harming the journalism business.
“The Times is the first major American media organization to sue the companies, the creators of ChatGPT and other popular A.I. platforms, over copyright issues associated with its written works,” the Times explained in a story that it published today, noting how it tried to work something out with Microsoft and OpenAI but realized that court action was necessary after “talks reached no resolution.”
The New York Times sued OpenAI and Microsoft for copyright infringement, a new front in the debate over the use of published work to train AI. https://t.co/u8qZ247dCl
— The New York Times (@nytimes) December 27, 2023
According to the complaint, a full copy of which can be found here, chatbots have been providing users with “near-verbatim excerpts from Times articles that would otherwise require a paid subscription to view,” with Microsoft and OpenAI said to be placing a “particular emphasis on the use of Times journalism in training their A.I. programs because of the perceived reliability and accuracy of the material.”
“In one example of how A.I. systems use The Times’s material, the suit showed that Browse With Bing, a Microsoft search feature powered by ChatGPT, reproduced almost verbatim results from Wirecutter, The Times’s product review site,” The Times noted elsewhere in its coverage.
Additionally, the problem of A.I. “Hallucinations” was brought up by the writers, with Microsoft’s Bing Chat alleged to have “provided incorrect information that was said to have come from The Times, including results for the 15 most heart-healthy foods,’ 12 of which were not mentioned in an article by the paper.”
“If The Times and other news organizations cannot produce and protect their independent journalism, there will be a vacuum that no computer or artificial intelligence can fill,” reads one portion of the complaint. “Less journalism will be produced, and the cost to society will be enormous.”
Susman Godfrey, the law firm that represented Dominion Voting Systems in its defamation case against Fox News, has been retained by The Times as its lead outside counsel.
Discussion (6 replies)
Join Discussion →Ah, too bad they used NY times, you know garbage in garbage out and all that jazz. Open AI should sue NY Times for harming their AI algorithm with their garbage and hubris.
By the same you can sue anyone referencing your articles in a research paper. Or even just for reading them and later relying on information learnt from it.
Headline from the future: NY times lays off 80% of "journalists" after deploying newsroom AI, a new news writing technology developed by Open AI
Technically, you are supposed to get permission before citing a copyrighted source, even in a non-profit or academic setting. An author or copyright holder can sue a researcher for copyright violations if they find fault with the use.
There is a "Fair Use" clause in US copyright law, and that allows "limited" use of copyrighted source, but there's no strict definition on what "Limited" means - just whatever you can justify in front of a judge and convince them of, if you get called to task.
Generally, unless someone is making money of someone else's copyright (which is the case here), or someone is using Fair Use to tarnish a brand or reputation - there's no real money in going after infringement, so you don't see a lot of enforcement. Unless you are Nintendo, then you just go after everyone.
If the NYT can prove that ChatGPT was intentionally and specifically trained using their material, or if they can catch ChatGPT using exact quotes or citations from NYT material without permission, they have a strong case and would probably at least settle something, if not outright win.
If it's just a case of ChatGPT scraping everything up in the web, and NYT had a lot of stuff hanging out there in the public domain - that's a less strong case, but has much wider implications. There are a lot of copyright holders out there with stuff hanging out, and it goes to show if scraping for commercial use is covered under Fair Use. It could impact non-AI things like search engines, which rely (in part) on crawling and scraping to generate metadata and indexes.
Imagine that you cite a famous author in your school paper, how on god green's earth would you get in contact with a world famous author to get their permission? And what do you mean by "fault with the use"? I doubt copyright law would use this vague terminology.
It is actually not that hard to prove fair use. If your work is transformative. And I doubt there is a more transformative use than generative AI.
For Nintendo it is not about fair use they go after anyone using their IP. Using an article in research or training is very different than releasing a supermario product.
Of course it was intentionally trained on it. How else? By accident? But not specifically only on that. They don't have a case even if someone trains a generative ai exclusively on NT articles, because the end product does not contain any of them. It's as if a teacher shows NYT articles to budding journalist in class then the NYT sues him for training them with their work.
They'd have a case if the generative model itself contained their copyrighted work, but it doesn't. If it was enough to sue someone for merely using copyrighted work as reference, then every journalist could be sued for reading other articles while researching for their own.
Contact their management or publisher I would assume