OpenAI Expands ChatGPT Memory for More Personalized Conversations

OpenAI is rolling out a new memory feature in ChatGPT that allows the AI to reference past conversations, creating more relevant and personalized responses.

The update, available to ChatGPT Plus and Pro users, enables the system to recall details such as preferences, names, and goals from earlier chats. It will apply to text, voice, and image interactions, streamlining communication by reducing the need for users to repeat information.

Initially, the feature is launching globally—excluding users in the U.K., European Economic Area (EEA), Switzerland, Norway, Iceland, and Liechtenstein, where OpenAI says additional regulatory reviews are needed. The company has committed to expanding availability in those regions in the future.

Key Features and User Controls

• More Capacity: ChatGPT can now remember a wider range of helpful information over time—such as preferences, workflows, recurring tasks, and personal details. Think of it as a smarter, more expansive journal that grows with your conversations.

• Smarter Organization: The memory system is better at connecting the dots across topics, even if they haven’t come up in a while. Whether it’s a recurring project, a specific name, or your favorite almond butter recipe, ChatGPT can now recall and apply these references more naturally across your chat history.

• Improved Accuracy: With the new memory update, ChatGPT is more consistent in how it applies your instructions—like tone preferences, formatting styles, or even recurring habits. This allows for a more seamless and reliable experience across different types of conversations.

• Smoother Updates: Users can update ChatGPT’s memory on the fly without starting from scratch. Simply say something like “Remember I prefer this phrasing,” and it will adapt in real time.

• Saved Memories & Chat History Reference: ChatGPT can store specific details for future use or reference past chats, even if information wasn’t formally saved. Saved memories are best for anything you want top-of-mind; chat history helps connect broader conversational context.

• Full User Control: You can view what ChatGPT remembers, delete specific memories, reset everything, or turn memory off entirely in Settings. For memory-free conversations, start a Temporary Chat that won't store or update information.

The feature is enabled by default for users who previously opted into memory capabilities, but users remain in full control. Opt-out options for both memory and chat history references are easily accessible in settings.

This new functionality builds on OpenAI’s previous memory tools, which required manual prompting to update or forget details. With this release, the process becomes automatic and seamless.

Availability and Access

• Eligible Users: ChatGPT Plus and Pro subscribers.

• Free Tier Access: Limited to Saved Memories only, without chat history reference.

• Geographic Limitations: Not currently available in EU, U.K., Switzerland, Iceland, Norway, or Liechtenstein. OpenAI states that these regions require further external reviews to meet local regulatory requirements, but the company remains committed to launching the feature there in the future.

The company notes that the more users interact with ChatGPT, the more useful it becomes—adapting and refining its responses over time.

What This Means

While memory in ChatGPT isn’t new, this update significantly expands what the feature can do—making it more capable, accurate, and personalized. Rather than simply storing a few user-defined preferences, ChatGPT can now remember and apply a broader range of information across interactions, including tone preferences, recurring topics, and personal habits.

This enhancement points to a larger trend in AI development: persistent, context-aware systems that adapt to individual users over time. As memory becomes more sophisticated, interactions with AI tools like ChatGPT start to feel more fluid, responsive, and tailored—reducing friction in workflows and boosting productivity.

At the same time, OpenAI continues to emphasize transparency and user control, offering tools to view, update, delete, or turn off memory entirely. This balance between personalization and privacy will be critical as memory features become more integral to how users interact with AI.

Looking ahead, these upgrades signal a move toward AI assistants that evolve with users—learning not just what you say, but how you work and what you value.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.