New York Times says OpenAI knew AI threatened publishers

In short: The New York Times claims in a new court filing that OpenAI and Microsoft copied millions of its articles to train AI systems, even while warning internally that this could seriously harm publishers.

What happened

The New York Times has filed new arguments and evidence in its copyright lawsuit against OpenAI and Microsoft. The case focuses on whether the companies broke copyright law by using publishers’ writing to train “large language models” (AI systems that learn patterns from huge amounts of text, like a student who reads a giant library to learn how to write).

In the filing, the Times says the defendants copied millions of copyrighted articles “in their entirety” without permission. The newspaper argues that this helped create paid AI products that can act as a substitute for reading the original articles.

The filing includes quotes the Times says come from internal messages and testimony. It says OpenAI’s president and co-founder Greg Brockman wrote around 2017 that he was “deeply motivated by the gazillions” he hoped to gain from the technology. It also claims OpenAI’s head of ChatGPT warned that publishers faced an “existential threat” from AI tools that are “largely substitutive” and could become more substitutive over time.

The Times also alleges OpenAI used “deceptive” methods to get paywalled content, and that Brockman responded “ah nice” after being told about “a hack to get around NY Times paywall.”

OpenAI has said training on publicly available content is “fair use,” similar to how search engines index web pages. Microsoft CEO Satya Nadella testified that he would have required OpenAI to retrain models if he had known paywalled data was used.

Why it matters

This lawsuit could shape how AI companies can use books, news, and other writing. If AI answers reduce visits to news sites, publishers could lose subscription and ad money, which can affect what news gets funded.

Source: Financial Times