April 11, 2025 – OpenAI has quietly rolled out an upgrade to its popular AI chatbot, ChatGPT, granting subscribers of Plus and Pro tiers access to an “enhanced contextual memory feature.”
CEO Sam Altman emphasized that the update now allows ChatGPT to “draw from a user’s entire conversation history,” enabling more tailored interactions. The system now adapts responses by analyzing past preferences and interests, whether gleaned from global usage data or explicitly stated by users via prompts.
This functionality empowers users to shape ChatGPT’s output by directly inputting their likes, dislikes, or topics of interest, streamlining the creation of personalized experiences. For example, a user could instruct ChatGPT to “always reference my interest in sustainable technology” or “avoid discussing political topics.”
Privacy controls remain robust, with users retaining the ability to edit or delete specific memories stored in their account settings. Alternatively, they can disable contextual memory entirely, reverting to a stateless interaction model. OpenAI positions the update as a balance between convenience and user agency, aiming to deepen engagement without compromising data stewardship.
The move reflects broader industry trends toward hyper-personalized AI, though critics warn of potential biases or over-reliance on historical data. OpenAI has yet to disclose technical specifics regarding memory retention limits or how it mitigates echo chamber effects.
