\n
<\/p>\n
OpenAI and Microsoft executives privately described their AI products as substitutes for journalism, according to court filings made public on Thursday that \u200bthe New York Times and other news outlets say gut the tech companies\u2019 fair-use defense in a closely watched copyright case.<\/p>\n
OpenAI and its largest financial backer, Microsoft, have argued in the Manhattan federal court case that their AI training on millions of newspaper articles is protected as fair use because it transforms copyrighted material into new content and does not compete with or replace the news organizations\u2019 journalism.<\/p>\n
Previously redacted portions of the filing made public on Thursday revealed quotes from OpenAI co-founder and president Greg Brockman, Microsoft CEO Satya Nadella and others during the discovery process that the news organizations say \u201ceviscerate\u201d the companies\u2019 \u200cfair-use defense to \u200cthe infringement claims.<\/p>\n
The outlets quoted OpenAI\u2019s head of ChatGPT, Nick \u200bTurley, \u200cas saying \u2060that \u200bpublishers face \u2060an \u201cexistential threat\u201d from their products, which are \u201clargely substitutive\u201d and \u201cwill get more and more substitutive as they get better.\u201d<\/p>\n
Spokespeople for OpenAI did not immediately respond to requests for comment on the filing.<\/p>\n
A New York Times spokesperson declined to comment.<\/p>\n
\u201cThe evidence revealed here for the first time shows that OpenAI and Microsoft knew that what they were doing was wrong,\u201d Steven Lieberman, an attorney for the New York Daily News and seven other newspapers involved in the case, said in a statement.<\/p>\n
The lawsuit, first filed in 2023, accuses OpenAI \u2060and Microsoft of using millions of articles without permission to train ChatGPT from \u200cnews sources that also include Ziff Davis publications, the Intercept \u200cand the Center for Investigative Reporting.<\/p>\n
Dozens of other complaints have also \u200bbeen brought by copyright owners against tech companies \u200cover what they call the theft of their material to train AI systems.<\/p>\n
The cases will likely \u200cturn on whether AI systems make fair use of copyrighted material by using it to create transformative new content. That question was at the center of arguments OpenAI and Microsoft made to US District Judge Sidney Stein in September 4 court filings.<\/p>\n
The Trump administration supports the AI companies and told Stein on September 1 that AI \u200ctraining is \u201cextraordinarily\u201d transformative.<\/p>\n
According to the outlets\u2019 Thursday filing, Brockman wrote that large language models are \u201c\u2019particularly good at predicting text of news articles,\u2019 \u2018excellent at news,\u2019 and \u2018very good \u2060at any news task,\u2019\u201d \u2060and responded \u201cah nice,\u201d when he was told that OpenAI employees had discovered a \u201chack\u201d to get around the New York Times\u2019 paywall.<\/p>\n
Nadella agreed under oath that conversing with chatbots \u201chas substituted … giving you the information right there on the website on the AI platform versus needing to go to the underlying source.\u201d<\/p>\n
Microsoft said Nadella\u2019s comment did not undermine the company\u2019s defense in the copyright case.<\/p>\n
\u201cSatya\u2019s testimony and Microsoft\u2019s position in this case are perfectly consistent,\u201d a Microsoft spokesperson said in a statement on Thursday. \u201cHe spoke to broad principles and changes underway in how people find and consume information.\u201d<\/p>\n
\u201cThose observations should not be confused with conclusions about copyright questions before the Court, which Microsoft addresses in its filings,\u201d the spokesperson added.<\/p>\n
The filing also quoted Microsoft\u2019s director of applied science, Brent Hecht, as saying that \u201cmillions of people\u201d would consider AI \u200bcompanies \u201c\u2019hoovering up\u2019 all their work\u201d to be \u201can astonishing \u200btheft of unprecedented proportions.\u201d<\/p>\n
\u201cThese comments reflect one employee\u2019s individual perspective, are not a legal analysis, and do not represent the company\u2019s views,\u201d the Microsoft spokesperson said in response.<\/p>\n<\/div>\n