NYT Copyright Brief Quotes Microsoft and OpenAI Leaders on AI Scraping Risks

Summary

The New York Times has asked for summary judgment in its late-2023 copyright lawsuit against OpenAI and Microsoft, arguing that the companies infringed copyright by using thousands of news articles. Its legal brief cites internal documents and testimony, much of which remains sealed or redacted at the companies’ request. A January 2023 memo by Microsoft Applied Science director Brent Hecht reportedly called large models “hoovering up” people’s work “the largest theft of labor in human history,” while another Microsoft document said creators generally had not authorized or been compensated for such use. Microsoft data cited in the brief reportedly showed that Copilot reduced click-through rates to the Times by as much as 93% compared with Bing search, prompting Hecht to describe a possible “doom loop” for both publishers and the web. OpenAI head of ChatGPT Nick Turley called ChatGPT an “existential threat” to publishers because its answers were increasingly substitutive, and an OpenAI engineer testified that prominent links still would not make users click. The filing also cites an internal exchange about bypassing the Times’ paywall. OpenAI and Microsoft maintain that internet scraping can qualify as fair use, and the article notes that one court found Anthropic’s use of published material covered by that doctrine. Microsoft CEO Satya Nadella nevertheless testified that paywalled content should be licensed and said Microsoft would have required retraining if it had known OpenAI trained on such material. The Times argues that these admissions could weaken the companies’ fair-use defense by showing awareness of the market effects of AI-generated answers.