Microsoft filed a lawsuit in federal court Monday that seeks to dismiss parts of a lawsuit brought by The New York Times Company.
The Times sued Microsoft and its partner OpenAI on December 27, accusing the two companies of infringing its copyright by using its articles to train artificial intelligence technologies such as the online chatbot ChatGPT. Chatbots compete with the news outlet as a source of reliable information, the lawsuit said.
In its filing, filed in the U.S. District Court for the Southern District of New York, Microsoft argued that large language models, or LLMs — the technologies that drive chatbots — are no substitute for buying news articles and other material they were trained on. .
The tech giant compared LLMs to videotapes, arguing that both are allowed under the law. “Despite the Times’ claims, copyright law is no more a barrier to the LLM than it is to video (or the player piano, duplicating machine, personal computer, Internet or search engine),” it said the movement.
In the late 1970s, movie studios sued Sony over its Betamax VCR, claiming it would allow people to illegally copy movies and TV shows. But the courts eventually found that making those copies for personal viewing was fair use under the law.
Microsoft’s move was similar to what OpenAI did last week. Microsoft said three parts of the lawsuit should be dismissed in part because the Times failed to show actual harm.
The Times had argued, for example, that if readers use Microsoft’s chatbot to research recommendations from the review site Wirecutter, which The Times owns, it loses revenue from users who would click on its referral links. Microsoft argued that the Times’ lawsuit offered “no real-world evidence to suggest significant diversion of revenue from Wirecutter.”
Ian Crosby, a Susman Godfrey partner who is the Times’ lead counsel in the case, said in a statement Monday: “Microsoft does not dispute that it worked with OpenAI to copy millions of Times works without its permission to build its tools . . Instead, he oddly compares LLMs to the VCR, even though VCR manufacturers have never argued that it was necessary to engage in massive copyright infringement to make their products.’
Microsoft had no immediate comment.
The Times was the first major US media company to sue Microsoft and OpenAI over copyright issues related to its written works. Writers, computer coders, and other groups have also filed copyright lawsuits against companies that make genetic artificial intelligence, technologies that generate text, images, and other media.
Like other AI companies, Microsoft and OpenAI built their technology by feeding it vast amounts of digital data, some of which is likely copyrighted. AI companies have claimed that they can legally use such material to train their systems without paying for it because it is public and they do not reproduce the material in its entirety.
In his suit, the Times included examples of OpenAI technology that reproduces excerpts from his articles almost verbatim. Microsoft said that training the technology in such articles was “fair use” under the law because chatbots were a “transformational” technology that created something new with copyrighted material. However, he did not seek to dismiss the arguments against “fair use”, saying he would address those issues later.