‘Doom Loop’: OpenAI and Microsoft Admit LLMs Is Destroying the Web and Built on Theft
microsoft openai
| Source: Mastodon | Original article
OpenAI and Microsoft acknowledge their large language models are damaging the web and are built on stolen content.
OpenAI and Microsoft have now put on the record what critics have long alleged: the large‑language models (LLMs) that power ChatGPT, Bing and other generative‑AI services are “predatory pieces of technology” built on what a senior Microsoft executive described as “an astonishing theft of unprecedented proportions” and “the largest theft of labor in human history.” The admission comes from internal documents that were unsealed in the New York Times lawsuit against the two firms. The papers warn that the companies’ data‑harvesting practices create a “doom loop” – AI systems scrape the web, generate content that replaces the original material, and then feed that output back into the training pipeline, eroding the very ecosystem that supplies their data.
The revelation matters because it sharpens the legal and regulatory spotlight on generative AI. If LLMs are indeed siphoning copyrighted text, images and other creative work without compensation, content creators and news organisations could face a rapid decline in traffic and revenue. The “doom loop” narrative also fuels ongoing policy debates in the EU and the United States about data‑rights, fair use and the need for transparency in AI training sets. It adds a new dimension to the Texas consumer‑safety suit we covered on 27 Oct 2026 and echoes the “largest theft of labor” comment we reported earlier this month.
Watchers should monitor the next steps in the New York Times case, any regulatory inquiries that may follow, and how OpenAI and Microsoft respond – whether by altering data‑collection practices, offering compensation mechanisms, or lobbying for new legal frameworks. The outcome could reshape the economics of the web and set precedents for how AI developers source and use online content.
Sources
Back to AIPULSEN