GitLab

OpenAI and the White House have implicated DeepSeek of using ChatGPT to inexpensively train its new chatbot.
- Experts in tech law state OpenAI has little recourse under copyright and agreement law.
- OpenAI's regards to use might use but are largely unenforceable, they say.
This week, OpenAI and the White House accused DeepSeek of something akin to theft.

In a flurry of press declarations, scientific-programs.science they stated the Chinese upstart had bombarded OpenAI's chatbots with questions and hoovered up the resulting data trove to quickly and inexpensively train a model that's now nearly as excellent.

The Trump administration's top AI czar said this training process, called "distilling," amounted to copyright theft. OpenAI, meanwhile, told Business Insider and other outlets that it's examining whether "DeepSeek may have inappropriately distilled our models."

OpenAI is not saying whether the business prepares to pursue legal action, instead assuring what a spokesperson called "aggressive, proactive countermeasures to safeguard our innovation."

But could it? Could it sue DeepSeek on "you took our material" grounds, similar to the premises OpenAI was itself sued on in an ongoing copyright claim filed in 2023 by The New York Times and other news outlets?

BI positioned this concern to experts in technology law, who stated challenging DeepSeek in the courts would be an uphill struggle for OpenAI now that the content-appropriation shoe is on the other foot.

OpenAI would have a difficult time proving a copyright or copyright claim, these attorneys stated.

"The question is whether ChatGPT outputs" - indicating the answers it generates in action to questions - "are copyrightable at all," Mason Kortz of Harvard Law School said.

That's since it's unclear whether the answers ChatGPT spits out qualify as "creativity," he said.

"There's a teaching that says creative expression is copyrightable, however facts and ideas are not," Kortz, who teaches at Harvard's Cyberlaw Clinic, stated.

"There's a big concern in copyright law today about whether the outputs of a generative AI can ever constitute innovative expression or if they are necessarily unguarded facts," he added.

Could OpenAI roll those dice anyhow and declare that its outputs are protected?

That's not likely, the lawyers said.

OpenAI is already on the record in The New York Times' copyright case arguing that training AI is an allowable "fair usage" exception to copyright defense.

If they do a 180 and inform DeepSeek that training is not a reasonable usage, "that might come back to kind of bite them," Kortz said. "DeepSeek could state, 'Hey, weren't you just stating that training is reasonable use?'"

There may be a difference between the Times and DeepSeek cases, Kortz added.

"Maybe it's more transformative to turn news articles into a model" - as the Times implicates OpenAI of doing - "than it is to turn outputs of a model into another model," as DeepSeek is said to have actually done, hikvisiondb.webcam Kortz said.

"But this still puts OpenAI in a pretty predicament with regard to the line it's been toeing regarding reasonable use," he included.

A breach-of-contract lawsuit is more most likely

A breach-of-contract claim is much likelier than an IP-based suit, though it comes with its own set of issues, stated Anupam Chander, who teaches technology law at Georgetown University.

Related stories

The regards to service for Big Tech chatbots like those developed by OpenAI and Anthropic forbid using their content as training fodder for a completing AI model.

"So possibly that's the claim you may potentially bring - a contract-based claim, not an IP-based claim," Chander stated.

"Not, 'You copied something from me,' however that you gained from my model to do something that you were not enabled to do under our contract."

There might be a hitch, Chander and Kortz stated. OpenAI's terms of service require that most claims be dealt with through arbitration, hb9lc.org not suits. There's an exception for suits "to stop unauthorized usage or abuse of the Services or copyright violation or misappropriation."

There's a larger hitch, though, specialists said.

"You must understand that the brilliant scholar Mark Lemley and a coauthor argue that AI regards to use are most likely unenforceable," Chander said. He was describing a January 10 paper, "The Mirage of Artificial Intelligence Regards To Use Restrictions," by Stanford Law's Mark A. Lemley and Peter Henderson of Princeton University's Center for Information Technology Policy.

To date, "no model creator has in fact tried to enforce these terms with financial charges or injunctive relief," the paper says.

"This is most likely for good reason: we think that the legal enforceability of these licenses is doubtful," it includes. That remains in part because design outputs "are mostly not copyrightable" and due to the fact that laws like the Digital Millennium Copyright Act and the Computer Fraud and Abuse Act "offer minimal recourse," it says.

"I believe they are most likely unenforceable," Lemley told BI of OpenAI's regards to service, "since DeepSeek didn't take anything copyrighted by OpenAI and because courts generally won't enforce agreements not to contend in the lack of an IP right that would avoid that competitors."

Lawsuits in between celebrations in various nations, each with its own legal and enforcement systems, are constantly difficult, Kortz said.

Even if OpenAI cleared all the above difficulties and won a judgment from an US court or arbitrator, "in order to get DeepSeek to turn over cash or stop doing what it's doing, the enforcement would come down to the Chinese legal system," he said.

Here, OpenAI would be at the mercy of another very complicated area of law - the enforcement of foreign judgments and the balancing of individual and business rights and national sovereignty - that stretches back to before the founding of the US.

"So this is, a long, made complex, stuffed procedure," Kortz added.

Could OpenAI have protected itself much better from a distilling attack?

"They might have utilized technical steps to block repeated access to their website," Lemley said. "But doing so would also hinder typical customers."

He added: "I do not think they could, or should, have a valid legal claim versus the searching of uncopyrightable details from a public site."

Representatives for DeepSeek did not immediately react to an ask for comment.

"We know that groups in the PRC are actively working to use methods, including what's called distillation, to attempt to reproduce advanced U.S. AI designs," Rhianna Donaldson, an OpenAI representative, informed BI in an .