AI Exec: We May Have Pulled Off “The Largest Theft of Labor in Human History”

AI search tools now send 90% fewer clicks to news articles.

Documents unsealed as part of lawsuits filed by Mother Jones and other publishers against OpenAI and Microsoft reveal internal discussions among tech executives regarding the use of copyrighted content for AI training. The lawsuits allege that these companies engaged in "astonishing theft of unprecedented proportions" by training AI models on vast amounts of text, including news articles, without compensation or consent from creators. Internal Microsoft documents reportedly acknowledged that "almost no one intended for content they created to be used in this fashion, nor are they compensated for its use." These documents also highlight concerns from executives about AI models potentially "substitut[ing] for the labor of the people that define the culture of society" and contributing to an "enshittification" of the internet by reducing clicks to original sources and overwhelming platforms with low-quality, AI-generated content. One executive noted that AI search resulted in a 90% reduction in clicks to news articles.

The unsealed documents expose a stark contrast between public statements and private acknowledgments from AI companies about their data acquisition practices. Executives appear to have recognized the ethical and economic implications of their training methods, including the potential to undermine the creators they rely upon and the overall quality of online information. This internal awareness, coupled with the documented "doom loop" concerns, suggests a corporate strategy that prioritized rapid development and market dominance over established rights and long-term ecosystem health.

The revelations indicate that AI companies may have deliberately stripped copyright and author information from scraped data, effectively obscuring the origins of their training material. Furthermore, the implementation of filters to block content from suing publishers, rather than all copyrighted material, suggests a strategy focused on mitigating legal risk rather than upholding intellectual property principles. This paints a picture of companies aware of their infringing activities, taking selective measures to avoid litigation while continuing to benefit from the very content at the heart of the disputes.