Internal Docs Undermine AI Fair Use Defense
According to The Decoder, a recent court filing by The New York Times and other media plaintiffs features internal emails and sworn testimony from OpenAI and Microsoft executives that challenge the tech companies’ fair use defense for AI training. Among the disclosures, Microsoft director of applied science Brent Hecht reportedly characterized the practice as an astonishing theft of unprecedented proportions.
The joint summary judgment brief highlights admissions from major tech figures regarding the impact of AI chatbots on web traffic. Microsoft CEO Satya Nadella confirmed under oath that chatbot conversations can replace visits to original publishing sources, while internal Microsoft documents warned of a self-reinforcing content doom loop. Furthermore, OpenAI head of ChatGPT Nick Turley acknowledged that AI products pose an existential threat to publishers by offering substitutive services.
The filings also detail internal discussions regarding paywalls and licensing. The documents indicate that OpenAI utilized restricted datasets and bypassed standard terms of service, with co-founder Greg Brockman acknowledging workarounds for publisher paywalls. In response to the filings, representatives for both Microsoft and OpenAI emphasized that individual employee statements do not constitute legal conclusions on copyright law, while litigation regarding billions in potential damages continues in US District Court.
Based on reporting by the-decoder.com.
