Two California ChatGPT users have filed a proposed class action against OpenAI in the U.S. District Court for the Northern District of California, accusing the company of routing real user conversations to outside contractors without proper disclosure. The practice falls under an internal program called Project Lily, where contractors read and score chatbot responses on a scale of one to seven as part of a reinforcement learning from human feedback (RLHF) process. ChatGPT has more than 900 million weekly users, many of whom share sensitive information such as medical symptoms, legal questions, or personal problems with the chatbot. The plaintiffs are demanding that OpenAI switch off the « Improve the model for everyone » setting by default, require explicit consent before any human reads a chat, and add clear warnings within the chat interface. OpenAI must respond to the lawsuit by October 13.
Source: Read the original article
