OpenAI Doesn’t Want to Train on New York Times Data, Altman Says

Artificial intelligence doesn’t need vast quantities of training data from publishers like The New York Times Co., according to OpenAI Chief Executive Officer Sam Altman, in a response to allegations his startup is poaching copyrighted material.

“There is this belief held by some people that you need all my training data and my training data is so valuable,” Altman said Tuesday at Bloomberg House at the World Economic Forum’s annual meeting in Davos. “Actually, that is generally not the case. We do not want to train on the New York Times data, for example.”

The ChatGPT maker is in the midst of a major push to secure access to news content after the Times last month sued the startup and Microsoft Corp., its biggest investor, for allegedly causing billions of dollars in damage over copyright infringement. Such partnerships are key to OpenAI’s future as it balances the need for timely, accurate data to develop its models with public scrutiny about where that data is sourced from.

OpenAI Chief Executive Officer Sam Altman and Anna Makanju, the company’s vice president of global affairs, speak with Bloomberg’s Brad Stone at Bloomberg House at Davos.

Source: Bloomberg

OpenAI is in talks with publishers including CNN, Fox Corp. and Time to license news content, Bloomberg reported last week, citing people familiar with the matter. The company’s chief of intellectual property and content previously told Bloomberg that the startup was in the middle of dozens of potential licensing deals.

Read More: OpenAI in Talks With CNN, Fox and Time to License Content

“What we want to do with publishers — if they want — is when one of our users says, what happened at Davos today, we’ll be able to say, here’s an article from Bloomberg, here’s an article from the New York Times,” Altman said. “Some people want to partner with us, some people don’t.”

Prior to the lawsuit, OpenAI and the Times had been in contact since April regarding licensing and failed to reach a deal. The AI company is also facing class action suits from writers including comedian Sarah Silverman, Game of Thrones author George R.R. Martin, and Pulitzer-winning author Michael Chabon.

Read More: OpenAI in Talks for New Funding at $100 Billion Valuation

Other publishers have been more amenable to cooperation. OpenAI signed an agreement with the Associated Press to access some of the news agency’s archives. It cut a three-year deal in December with Axel Springer SE to use the German media company’s work for an undisclosed sum.

“A lot of our research is how do we learn more from smaller amounts of very high quality data,” Altman said.

To contact the reporters on this story:
Brad Stone in Davos at bstone12@bloomberg.net;
Jake Rudnitsky in Berlin at jrudnitsky@bloomberg.net

To contact the editors responsible for this story:
Lynn Doan at ldoan6@bloomberg.net

Graham Starr

© 2024 Bloomberg L.P. All rights reserved. Used with permission.