Unsealed documents reveal damning admissions

New unredacted filings in The New York Times' three-year-old copyright lawsuit against OpenAI and Microsoft show executives privately called AI scraping 'theft' and admitted the technology poses an 'existential threat' to publishers.

Microsoft director called it 'unprecedented theft'

In a January 2023 internal memo, Microsoft's director of Applied Science Brent Hecht described the scraping as 'an astonishing theft of unprecedented proportions' and 'the largest theft of labor in human history.'

“the largest theft of labor in human history”

Copilot tanked New York Times traffic

Microsoft's own data showed its Copilot answer engine caused click-through rates for The New York Times domain to drop as much as 93% compared to traditional Bing search results.

Microsoft warned of a 'doom loop'

An internal Microsoft presentation from January 2024 described the traffic decline as a 'doom loop' that would 'hurt the performance of our models and the entire web at the same time.'

Nadella said paywalled content needs licensing

Microsoft CEO Satya Nadella testified that 'anything that is paywalled should be licensed by anyone who wants to use it for grounding or training,' and said he would have required OpenAI to retrain its models if he'd known about the scraping.

“anything that is paywalled should be licensed by anyone who wants to use it”

OpenAI called ChatGPT 'substitutive'

OpenAI's head of ChatGPT Nick Turley wrote that publishers face an 'existential threat' from products that are 'largely substitutive' and 'will get more and more substitutive as they get better.'

“existential threat”

The scale of copying was massive

OpenAI's training datasets contained over 91,692 copies of works from The New York Times, Daily News, and Center for Investigative Reporting. A Common Crawl dataset included more than 2 million documents from nytimes.com alone.

They bypassed paywalls deliberately

OpenAI employees devised a plan to circumvent paywalls without detection. When researcher Nick Ryder told Greg Brockman about a 'hack to get around nytimes paywall,' Brockman replied 'ah nice.'

“ah nice”

They stripped copyright notices from data

OpenAI researchers deliberately removed copyright notices from training data before it reached the model, since they 'wouldn't want model outputting copyright notices to users.'

Fair use defense now looks weaker

These admissions contradict OpenAI and Microsoft's legal argument that training constitutes fair use, which requires that use doesn't harm the market for the original work.

AI could disrupt the creators themselves

A Microsoft document warned there is a 'real risk' that generative AI could 'significantly disrupt the employment of the very people who generated the data on which the foundation model was trained.'