Microsoft Execs Privately Called Web Scraping 'Largest Theft of Labor in Human History'
Unsealed court documents have revealed that Microsoft executives privately described web scraping as 'the largest theft of labor in human history.' Dr. Brent Hecht, Microsoft's Director of Applied Science, wrote in an internal memo that generative AI risks creating a self-destructive 'doom loop' by wiping out the economic foundations of its training data suppliers.
According to the court materials, OpenAI and Microsoft used methods to assemble massive training datasets, including bypassing paywalls. An internal message showed an OpenAI researcher informing company president Greg Brockman about a 'hack' to pull content from The New York Times without permission. Brockman responded with a casual 'ah, nice.'
The filings also reveal that researchers intentionally removed copyright notices from models to prevent users from seeing them. Internal metrics recorded click-through rates to news sites dropping between 83% and 93% when users interacted with Copilot instead of traditional search engines.
OpenAI's head of ChatGPT, Nick Turley, noted in internal communications that chatbots pose an 'existential threat' to publishers. He said chatbots provide little incentive for readers to click through to original sources.