OpenAI is in a “tough position” as it struggles to avoid a flood of copyright lawsuits from media companies, experts told the Post.
OpenAI's ChatGPT and other AI chatbots are allegedly exploiting copyrighted news content to compete for attention with media companies without giving proper credit or compensation to publishers. Experts say that if the fraud goes unchecked, it could have a devastating impact on traditional media businesses.
In an effort to keep the peace, OpenAI (whose CEO, Sam Altman, recently returned to management after a brief coup by the company's board of directors) has joined several prominent companies, including CNN and Fox Corp. It is reportedly in talks with media companies about deals. Pay for access to content you can use to train popular chatbots.
OpenAI has already announced content deals with Associated Press and Axel Springer.
But the company's disruptive AI tools, including its popular chatbots, have led to lawsuits like the one recently filed by the New York Times, which experts say could give OpenAI the ability to plagiarize Gray Lady's content. He said there was a frightening debate to either quit or collect hefty fees.
OpenAI's response so far has yielded mixed results. Experts say the company's piecemeal approach to negotiations risks putting it in trouble as U.S. lawmakers and federal courts alike examine the legality of AI training.
“They are in a tough position,” an industry source who requested anonymity to discuss the situation told the Post. “I think they're looking at what's going to happen now. If you negotiate with the collectives individually, each one will take different actions. Whatever happens from there, it's not that predictable. there is no.”
The New York Times said it chose to sue OpenAI only after talks about a possible “amicable solution” such as a licensing agreement broke down several months ago. Separately, The Washington Post has not held any negotiations in recent months, a spokesperson for the company confirmed to the paper.
The Times' lawsuit, filed last month in Manhattan District Court, seeks to hold OpenAI and lead backer Microsoft liable for “billions of dollars” in damages. It will likely be many years before a lawsuit begins, and companies face the risk of their work being stolen in the meantime, with no economic or legal repercussions.
The lawsuit includes numerous examples of GPT-powered chatbots spitting out verbatim or near-exact copies of Times articles in response to user prompts, including one in 2012 when a celebrity Including the infamous scathing review of Chief Guy Fieri's American Kitchen & Bar restaurant and the Pulitzer. Award-winning article “Snow Fall: The Avalanche at Tunnel Creek.”
In a lengthy blog post on January 8, OpenAI called the Times' lawsuit “without merit” and said the paper “deliberately uses prompts that often include long excerpts of articles to regurgitate our model.” “It was manipulated,” he claimed.
The blog post cites “long and widely accepted precedent,” but lacks specific examples of “fair use” litigation to support its position, making OpenAI appear “a little scared.” said an industry official.
“If you're on the weak side, that's when the company should just sit back and wait for the court and hope the judge isn't afraid to break the shiny toy,” the source added.
OpenAI dismissed the “regurgitation” of verbatim passages from the article as “a rare bug and we are working to eliminate it.” The company said the Times refused to provide the case before the lawsuit, which would have been quickly corrected.
James Grimmelman, a professor of digital and information law at Cornell University School of Law, said the Times' lawsuit is against AI companies because the paper “submitted receipts” showing specific examples of near-perfect copy. “This is the most serious lawsuit ever filed.”
“Many other lawsuits rely on much thinner copy representations, such as book summaries or showing that they can be generated one sentence at a time,” Grimmelman told the Post. “This indicates that ChatGPT remembers a large number of Times articles.”
Cristeria Garcia, a copyright law expert and professor at Georgetown University Law Center, said it would be surprising if OpenAI were allowed to sue, given the huge financial and legal stakes involved in a lawsuit. .
A potential court ruling that could upend the entire business model could put pressure on OpenAI to settle or accept a retroactive license agreement to resolve the Times' claims. There is sex.
“The statutory damages are enormous,” Garcia added. “Effectively, it would not only shut down the business model as they know it, but it would probably shut down the company.”
The Post has reached out to OpenAI for comment.
Elsewhere, an anonymous media company is reportedly “considering legal action” against OpenAI, similar to the Times lawsuit. reported by bloomberg. Last year, billionaire investor Barry Diller suggested that publishers: AI companies should be sued for misusing content.
The debate spilled over to Capitol Hill, where Condé Nast CEO Roger Lynch told a Senate committee that AI tools were “made with stolen goods” and called on Congress to enact regulations. .
OpenAI also reportedly irritated some media executives with lackluster offers. Last week, The Information reported that OpenAI offered news organizations less than $5 million in exchange for licenses to use articles.
Despite ongoing negotiations with other media outlets, OpenAI is clearly bracing for an onslaught of legal action related to its business practices.
Eyebrow-raising confessions in filings OpenAI told the House of Lords last week that it is “impossible to train today's leading AI models without using copyrighted material”.
Meanwhile, the Times and other news organizations that choose to litigate may consider settling, Grimmelman said. This is because the agreement for periodic license fee payments would provide the paper with a “long-term sustainable model” to profit from the use of AI work.
