OpenAI has released a new update to ChatGPT, giving the AI chatbot the ability to remember past conversations.
This marks a significant change, allowing the chatbot to use previous interactions when generating responses across text, voice and image inputs.
OpenAI’s CEO Sam Altman said the update is designed to create a more personal AI assistant that learns and grows with users.
“We have greatly improved memory in ChatGPT – it can now reference all your past conversations,” Altman said on X.
He called the feature “surprisingly great” and believes it shows the future of AI: tools that understand users more deeply.
At present, the memory feature is only available to paying subscribers of ChatGPT, with Pro users receiving immediate access.
Plus users will get the update soon, but free-tier users will not have access to memory for now, OpenAI confirmed.
Users in the UK, EU, Iceland, Liechtenstein, Norway and Switzerland will also need to wait due to legal requirements.
Importantly, users can opt out of memory or use a “Temporary Chat” option that prevents any data from being saved.
The launch follows a similar move by Google, which added memory to its Gemini AI chatbot earlier this month.
However, researchers recently exposed a flaw in Gemini, showing that hackers could insert false memories into the AI.
Google has responded quickly, fixing the issue and strengthening safeguards around its AI’s long-term memory function.
OpenAI’s update is another step toward creating AI that feels more personal, useful and truly built for individual users.
