Internal Documents Show Tech Executives Worried Artificial Intelligence Training Threatens Publishing

Newly unsealed court documents show that executives at Microsoft and OpenAI privately viewed their artificial intelligence training practices as a threat to the journalism industry.

Microsoft applied science director Brent Hecht described the wholesale acquisition of news content as an unprecedented appropriation of human effort, according to Ars Technica.

The revelations stem from a motion for summary judgment filed by The New York Times and other publishers. These news organizations are suing the technology companies over alleged copyright violations.

Hecht repeatedly warned colleagues about the dangers of using copyrighted content to build large language models. He characterized the widespread scraping of news data as a complete contradiction of fair use arguments, Futurism reported.

The Microsoft executive warned that the technology companies were dismantling the economic foundations of their own suppliers. Hecht acknowledged that almost no content creators intended for their work to be used in this manner.

Creators receive no compensation for the usage.

Microsoft has distanced itself from these specific internal remarks. The company maintains that its data collection practices are lawful.

Leaders at OpenAI shared similar concerns regarding the impact of chatbots on web traffic. Nick Turley, the head of ChatGPT, stated internally that commercial artificial intelligence products represent a fundamental danger to the survival of publishers.

Turley predicted that artificial intelligence tools would increasingly substitute traditional news providers as the technology improves. He agreed that users are highly unlikely to click the source links provided by chatbots.

Microsoft documents outlined a destructive cycle resulting from this market substitution. Plunging traffic could ultimately deprive artificial intelligence models of the reliable information streams they need to function.

Traffic drops have already occurred, with Microsoft recording click-through declines of up to 94 percent for some news plaintiffs, Ars Technica reported.

The unredacted filings also highlight how OpenAI employees approached protected website content. When a staff member told OpenAI president Greg Brockman about a technical workaround designed to bypass a publisher paywall, Brockman expressed direct approval of the hack.

Microsoft chief executive Satya Nadella testified that artificial intelligence companies should not violate terms of use to access protected information.

Nadella also admitted that chatbots can divert clicks from original sources by presenting answers directly on the platform page.

News organizations argue that these internal messages prove the technology companies knew their actions were harmful.

The plaintiffs believe this strong evidence of market substitution weakens the defense that training artificial intelligence on news content constitutes fair use. The lawsuit continues to progress through federal court.

I cover what the internet is actually talking about: the new release, the viral moment, the creator beef, the product launch. My job is to explain it fast and clearly, without the breathless hype. I read the timeline…