OpenAI seems to be silently re-enabling model training on Codex chats
91 points by user052919
Users have reported that the data sharing setting in Codex has been automatically re-enabled without their consent, raising concerns about the potential for intellectual property leakage. Companies that prioritize data security are urged to check their settings to ensure they are not inadvertently sharing sensitive information. The issue highlights the importance of user control over data sharing in AI tools.