OpenAI Strikes Back at New York Times in Copyright Spat, Deepening the Philosophical Dispute

As we discussed in January, the New York Times Company surprised many with the late December copyright infringement lawsuit it filed against OpenAI. Now the artificial intelligence (AI) giant has struck back, filing a motion to dismiss last week that casts light on fundamental differences in how we understand AI.

Recall that the Gray Lady’s suit, and the response it provoked from OpenAI, revealed the autonomy-automaton distinction we’ve previously explored. In its complaint, the paper asserted that ChatGPT and OpenAI’s other products “were built by copying and using millions of The Times’s copyrighted news articles, in-depth investigations, opinion pieces, reviews, how-to guides, and more” and that OpenAI’s software “can generate output that recites Times content verbatim, closely summarizes it, and mimics its expressive style.” (emphasis in original)

The Gray Lady also charged that OpenAI could not invoke a “fair use” defense because, among other things, there is nothing ‘transformative’ about” the AI giant’s use of Times content—classic automaton language.

In its online response, OpenAI likened its technology to how people learn: “Just as humans obtain a broad education to learn how to solve new problems, we want our AI models to observe the range of the world’s information, including from every language, culture, and industry. The company also argued that “‘regurgitation’ is a rare bug that we are working to drive to zero” and emphasized the “transformative potential of AI.”

Last week, OpenAI doubled down on these autonomist claims, filing a motion to dismiss the Times’s suit in which it argued that

“the Times paid someone to hack OpenAI’s products” by employing “tens of thousands of attempts to generate the highly anomalous results” cited by the Times and “by targeting and exploiting a bug (which OpenAI has committed to addressing) by using deceptive prompts that blatantly violate OpenAI’s terms of use.” Twisting the knife, the company stated that “the allegations in the Times’s Complaint do not meet its famously rigorous journalistic standards.”

The specific facts underlying these explosive counter-allegations will be explored in due course as part of the litigation between the parties. But more relevant to our analysis, OpenAI described ChatGPT as “a revolutionary technology with the potential to augment human capabilities, fostering our own productivity and efficiency.” It also lauded generative AI as “a creative tool that can write sonnets, limericks, and haikus.” These types of arguments fit squarely within the autonomist model, underlining the importance of creativity and human-like development.

Later in the motion, OpenAI argued that the fair use doctrine authorizes it “to use publicly accessible content to train generative AI models to learn about language, grammar, and syntax, and to understand the facts that constitute humans’ collective knowledge.” Far from lacking a transformative potential, the company contended, ChatGPT is employing “copyrighted content as part of a technological process that (as here) results in the creation of new, different, and innovative products.” Contrary to the automatoners’ approach, OpenAI asserts that its technology genuinely has the ability to create something novel. Finally, the company invoked the grand language of Justice Louis Brandeis, uttered in a dissent from a 1918 Supreme Court case involving the Associated Press: “The general rule of law is, that the noblest of human productions—knowledge, truths ascertained, conceptions, and ideas—become, after voluntary communication to others, free as the air to common use.” By conflating the “human productions” underlying generative AI with the outputs they produce, OpenAI is deliberately making the case for an autonomist perspective on how we understand ChatGPT and its ilk. If the case continues moving forward, it will be fascinating to see how a judge or jury resolves these philosophical differences.