Posted 2/28/2024, 1:45:00 AM
OpenAI Accuses New York Times of Exploiting ChatGPT to Generate Verbatim Passages; NYT Calls Hacking Claims Bizarre
- OpenAI accused the New York Times of paying someone to "hack" ChatGPT to generate verbatim passages from NYT articles
- OpenAI claimed the NYT abused ChatGPT by tricking it to regurgitate training data, exploiting a bug with deceptive prompts
- The NYT lawyer called OpenAI's hacking allegations "bizarre" and said they simply used ChatGPT to find evidence of copyright infringement
- OpenAI appears to monitor user prompts and responses, despite previously claiming not to track queries and outputs
- The NYT lawyer challenged OpenAI's argument that the complaint was brought too late, saying OpenAI concealed how its products operate