OpenAI and Microsoft Knew in Advance About the 'Doom Loop' for the Web

In the lawsuit against OpenAI and Microsoft by The New York Times, 92 pages of internal documents have been declassified: Microsoft's Director of Applied Sciences, Brent Hecht, called the collection of training data 'the largest labor theft in human history,' and Satya Nadella confirmed under oath that chatbots are replacing transitions to source sites.

Compiled by Sergey KostenchukPublished 2026-09-20Updated 2026-09-20

2026-09-20 Business OpenAI

💼 OpenAI and Microsoft Knew in Advance About the 'Doom Loop' for the Web

According to the lawsuit by The New York Times, 92 pages of internal documents have been declassified: Microsoft's Brent Hecht called the collection of training data 'the largest labor theft in human history,' and Satya Nadella confirmed under oath that chatbots are replacing transitions to publisher sites.

🌍 The companies have for the first time acknowledged in writing that LLMs 'destroy their own content supply chain': without clicks, publishers lose revenue and write less new text for training models. The outcome of the lawsuit will determine whether training violates copyright under fair use.

👤 The original source shows that top managers knew in advance about 'Google Zero' and the verbatim reproduction of articles from The Times, Mercury News, and The Denver Post, and at OpenAI, as early as 2022, they wrote that GPT-4 would 'remember a ton of data.'

Source 1: https://www.theverge.com/ai-artificial-intelligence/997633/openai-microsoft-chatgpt-ai-new-york-times-doom-loop-theft-google-zero

Sources