6 min readApr 6, 2024 09:25 AM IST
First published on: Apr 6, 2024 at 08:30 AM IST
Artificial intelligence (AI) is reconfiguring our lives and forcing us to question whether only human beings are capable of sophisticated intelligence? After all, AI is passing lawyers’ bar exams and other professional tests. It’s even performing precise but finite surgical procedures. However, there are some fetters in the AI revolution. And this has come from an unlikely quarter — copyright law. Novelists, newspapers, writers are amongst plaintiffs who have sued various companies using AI alleging that AI “ingests” their work. These plaintiffs from various suits include the popular author John Grisham and The New York Times.
The New York Times (NYT) in a very high-profile case, in late December 2023 sued OpenAI and Microsoft for copyright infringement. In their petition, the NYT alleges that the “defendants’ generative artificial intelligence (GenAI) tools rely on large-language models (‘LLMs’) that were built by copying and using millions of Times copyrighted news articles, in depth investigations, opinion pieces, reviews, how-to-guides, and more. While the defendants engaged in wide-scale copying from many sources, they gave Times content particular emphasis when building their LLMs — revealing a preference that recognises the values of those works.”
The NYT in these absolutely delightful pleadings, argues that the Constitution and the Copyright Act recognise the “critical importance of giving creators exclusive rights over their works”. The newspaper felt strongly that their copyright was violated by OpenAI and Microsoft and hence they were unable to enjoy the “fruits of their labour and investment.” How does the NYT explain this alleged exploitation of their labour? They argue that “powered by LLMs containing copies of Times content, Defendants Gen AI tools can generate output that recites Times content verbatim, closely summarises it, and mimics its expressive style”, as demonstrated by scores of examples. “Defendants also use Microsoft’s Bing search index, which copies and categorises the Times articles that are significantly longer and more detailed than those returned by traditional search engines.”
Essentially, the NYT is arguing that the defendants are using its intellectual property without paying for it and enriching themselves. The NYT links this use of its intellectual property protected work to the growth in Microsoft’s valuation to a trillion dollars, and OpenAI’s ChatGPT’s valuation of $90 billion.
In a bid to address the argument of “fair use” that is often a defence against copyright infringement, the NYT argued that there was nothing transformative about using NYT content without payment to create products that substitute for the newspaper and steal audiences away from it. The core argument of the NYT is that the outputs of “Defendants’ GenAI models compete with and closely mimic the inputs used to train them” copying NYT works and hence is not fair use.
The NYT’s petition includes extracts from searches conducted on ChatGPT, and Bing Chat (Microsofts’ GenerativeAI) that are alleged to throw up NYT articles verbatim.
Importantly, the petition for the NYT, while arguing that it has lost billions of dollars, does not ask for a specific amount for damages. It does ask for an injunction to stop the alleged unlawful conduct from continuing.
The reply of OpenAI is clever. In fact, the legal petitions filed by the three parties are rigorous and deeply enjoyable reading. For students of law (by that I mean those in law school and those practising law) the petitions are worthwhile to review for they are indicative of what good legal drafting ought to be. Succinct, clever and elegantly incorporating the law.
However, the case of The New York Times Co v Microsoft Corp et al is also fascinating since it has all the attributes of an international caper. How, you might ask? OpenAI responds to the allegations by arguing that “the Times paid someone to hack OpenAI’s products.” Further, that it took the newspaper ten of thousands of attempts to generate the “highly anomalous results” that constitute the illustrations in the complaint. Further, that they were able to do so only “by targeting and exploiting a bug by using deceptive prompts that blatantly violate OpenAI’s terms of use. And even then they had to feed the tool portions of the very articles they sought to elicit verbatim passages of, virtually all of which already appear on multiple public websites.”
Microsoft responded strongly by comparing the New York Times lawsuit to the one waged by the Motion Picture Association of America and Hollywood against the VCR (Video Cassette Recorder). When the VCR was first introduced, the entertainment industry claimed that it violated copyright. As Microsoft writes, the “(US Supreme) Court rejected the alarmism and voted for technological innovation and consumer choice, in its seminal decision in Sony Corp of America vs Universal City Studios, Inc. freeing consumers from broadcast schedules and ushering in an on-demand world.” Clearly, that decision did not destroy Hollywood — quite the opposite, it’s argued that the entertainment industry flourished “when the VCR opened new markets and revenue streams”.
News conglomerates have not all chosen to fight the use of their work product to feed the insatiable appetite of AI. Some have joined hands with it. The large European news conglomerate Axel Springer has announced a partnership with OpenAI to “strengthen independent journalism in the age of AI”. In July 2023 OpenAI and the Associated Press (AP) announced a deal for the AI company to license AP’s archive of news stories.
As The New York Times Co v Microsoft Corp et al weaves its way through the court system, more such challenges for the law will emerge as AI swiftly takes over how humans access, process and pay for news and creative work. Alternatively, we can ask ChatGPT — “are you using New York Times stories and reports without paying for it?” I suspect we will get a legally nuanced answer involving “fair use”.
The writer is a Senior Advocate at the Supreme Court
The NYT’s petition includes extracts from searches conducted on ChatGPT, and Bing Chat (Microsofts’ GenerativeAI) that are alleged to throw up NYT articles verbatim. (Illustration by CR Sasikumar) 
Share
