Is it just me or does OpenAI keeps enabling memory in a clandestine way
A user reports that OpenAI appears to have silently re-enabled the memory feature on their account, despite having previously disabled it. After logging in, the user dismissed a customization notification only to later find the memory flag toggled on in Settings.
Background
- OpenAI's "memory" feature lets ChatGPT retain personal facts and preferences across conversations (e.g., your name, dietary restrictions, or that you prefer short replies). Users can disable it in Settings, and OpenAI says it only learns from chats when memory is on.
- This user reports that their memory toggle was silently re-enabled after an update, despite having previously turned it off. They suspect OpenAI is enabling it without clear consent.
- The complaint touches on a broader trust/surveillance concern in the AI industry: companies have financial incentives to collect more user data to improve models and personalize advertising, even when they promise otherwise.
- Similar accusations have been leveled at other tech giants (Google, Meta) for toggling privacy settings back on after updates, making this a familiar pattern for privacy-conscious readers.