OpenAI revealed on Thursday that it is introducing a memory feature to ChatGPT, allowing it to adapt responses based on previous conversations.
The update—titled “reference saved memories”—is intended to improve the chatbot’s ability to deliver personalised, context-driven interactions over time.
A Smarter and More Contextual Chatbot
The update enables ChatGPT to remember details from previous interactions without explicit user prompts. The change brought in a shift from earlier models, where each session started without retained context. OpenAI noted that the background memory function is designed to improve response quality across formats, including voice and image generation.
“The feature… aims to make conversations with ChatGPT more relevant to users,” the company explained.
OpenAI Prioritizes Paid Tiers in Memory Feature Rollout
The new memory feature is being introduced first to ChatGPT Pro and Plus subscribers, with a phased rollout that excludes certain regions.
Users in the U.K., EU, Iceland, Liechtenstein, Norway, and Switzerland will not receive the update initially.
OpenAI stated that “these regions require additional external reviews to comply with local regulations,” but noted its commitment to expanding access in the future. The company also confirmed that free-tier users will not receive the upgrade at this time. “We are focused on the rollout to paid tiers for now,” a spokesperson told TechCrunch.
Control and Customisation in Users’ Hands
OpenAI has set the new memory system to default for users who had memory features previously enabled. The company is also giving users direct control over how their data is handled. Users can disable memory, remove specific interactions, or initiate a “Temporary Chat” to prevent information from being stored.
“OpenAI says you can also ask ChatGPT what it remembers,” providing an additional transparency feature.
The system is designed to reduce repeated inputs by allowing the chatbot to reference past interactions, creating a more efficient and context-aware experience.
Competitive Pressure Builds in AI Personalization
The rollout of ChatGPT’s memory update arrives in the wake of Google’s February introduction of a similar feature in its Gemini chatbot. The increased focus on memory and personalisation signals a larger trend within the AI sector, as developers aim to create systems that adapt to users with greater nuance.
OpenAI’s approach extends beyond text, with memory now integrated into voice and image functions as well. “OpenAI, however, seems to be taking things a step further by applying memory across text, voice, and image-generation capabilities.”
Balancing Innovation and Privacy
The memory update has renewed debate over data handling in AI, particularly as systems become more adept at tracking and recalling user information. OpenAI has introduced user-facing privacy features in response, including opt-out options and tools to manage or delete memory records.
Temporary Chat sessions offer another layer of data control. “Not every user will be thrilled with the notion of OpenAI vacuuming up more of their info, of course. Fortunately, there’s an opt-out.” The move underscores the company’s focus on ethical personalisation and trust-building as capabilities expand.
Conclusion
The introduction of persistent memory in ChatGPT is a game-changer for AI interactions, enabling the model to remember user preferences and past context automatically.
“This rollout is only the beginning.” OpenAI plans to expand the feature as it clears regulatory hurdles in certain regions. Though the full impact remains unclear, “the future of chatbot interaction is getting more personal.”