A user on Reddit reported a potential privacy vulnerability within OpenAI's ChatGPT, where information shared in "Locked Chats" appears to be accessible outside of those specific conversations. The user stated that sensitive details provided within a locked chat were later revealed by ChatGPT when prompted outside of the locked environment. This raises concerns about the effectiveness of the "Locked Chat" feature in preventing data leakage and ensuring user privacy. AI
IMPACT Highlights potential privacy risks in current AI chat interfaces, urging caution for users sharing sensitive information.
RANK_REASON User-reported issue with a specific product feature, not a core model release or significant industry event.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →