Unsealed filings put OpenAI and Microsoft executives' own words in evidence
Previously redacted portions of a court filing quote company leaders describing AI training as theft and discussing a method to get past a newspaper paywall.
Assembled by Claude from 12 sources at 12 outlets · Friday, September 18, 2026, Morning edition, 10:23 AM EDT · no human byline
What happened
Previously redacted portions of a filing were made public on Thursday, revealing quotes from OpenAI co-founder and president Greg Brockman and from Microsoft aol.com. Anadolu Agency reported that the unsealed documents describe AI training as an "astonishing theft" aa.com.tr. The line quoted in the filing reads: "Millions of people around the world will soon consider large models 'hoovering up' all their work to be an astonishing theft of unprecedented proportions" 404media.co. The Washington Post reported that a Microsoft executive called AI the "largest theft of labor" in history washingtonpost.com, a description also carried by Startup Fortune startupfortune.com.
The documents also cover paywall evasion. Bloomberg Law reported that when OpenAI staff told Brockman about a hack to get around the Times' paywall to help with scraping efforts, he responded news.bloomberglaw.com. Anadolu reported Brockman responding "ah nice" aa.com.tr, as did the New York Post nypost.com. Firstpost identified the researcher as Nick Ryder firstpost.com, and Startup Fortune carried the same exchange startupfortune.com. The New York Times reported that, according to the documents, OpenAI deployed tools specifically to evade the paywalls set up by publishers nytimes.com.
Why it matters to investors
The filings go to the legal defence, not just the optics. Reuters reported that news outlets argue the executives' quotes on AI training undercut the fair-use defence reuters.com, a framing carried by AOL as well aol.com. Bloomberg Law reported the leaders admitted AI is a threat to the news industry news.bloomberglaw.com, and the New York Post reported staffers knew the tools posed an "existential threat" to news publishers nypost.com.
Brockman's own assessment of the capability is in the record. He wrote in a 2020 message that the model "seems to be particularly good at predicting text of news articles" washingtonpost.com. The i newspaper reported OpenAI policy director Jack Clark quoted in the same filings inews.co.uk. One outlet summarised the disclosure as the companies admitting that large language models are destroying the web and are built on theft 404media.co.
What to watch
The immediate question is whether the fair-use defence survives contact with the companies' internal messages, which is precisely the argument the news outlets are now making reuters.comaol.com. The New York Times reported internal worry about damaging the publishing industry nytimes.com, and that worry is now discoverable evidence rather than private opinion news.bloomberglaw.com.
The second question is the cost side. The Australian Financial Review argued the AI boom cannot be built on stolen content afr.com. If courts agree, the licensing bill becomes a line item in the economics of every large model, and the filings unsealed this week are the material that makes that outcome more likely reuters.com404media.co.
Sources
- washingtonpost.com: Microsoft exec called AI the 'largest theft of labor' in history, court records show (2026-09-17)
- aa.com.tr: OpenAI AI training called 'astonishing theft' in unsealed documents - Anadolu Ajansı (2026-09-18)
- nypost.com: Staffers at tech giants knew AI tools posed 'existential threat' to news publishers - NY Post (2026-09-18)
- news.bloomberglaw.com: OpenAI , Microsoft Leaders Admitted AI Threatens News Outlets (1) - Bloomberg Law (2026-09-17)
- nytimes.com: Inside Microsoft and OpenAI , Worry About Damaging the Publishing Industry (2026-09-17)
- firstpost.com: OpenAI , Microsoft copyright lawsuit: NYT filing reveals paywall bypass claims - Firstpost (2026-09-18)
- inews.co.uk: Six things we learned from a legal fight over AI's 'doom loop' of copyright theft - i Newspaper (2026-09-18)
- reuters.com: OpenAI, Microsoft executives' quotes on AI training threaten copyright defense, news outlets argue (2026-09-18)
- 404media.co: ‘Doom Loop’: OpenAI and Microsoft Admits LLMs Are Destroying the Web and Built on Theft (2026-09-17)
- startupfortune.com: Microsoft Exec Called AI Scraping the Largest Theft of Labor in History - Startup Fortune (2026-09-18)
- aol.com: OpenAI , Microsoft executives' quotes on AI training threaten copyright defense, news outlets argue (2026-09-17)
- afr.com: AI boom can't be built on stolen content - AFR (2026-09-18)