Photo by Rodolfo Clix on Pexels

Anthropic is boosting its Claude AI chatbot with a new ‘memory’ feature for all paid subscribers, bringing it closer in functionality to rivals like ChatGPT and Gemini. This upgrade allows Claude to retain details from past conversations without being explicitly told to, making it more efficient and personalized. The feature, already available for Team and Enterprise users, is now rolling out to Max subscribers and will soon be available for Pro users. Anthropic prioritizes user control and transparency, allowing users to view, edit, toggle, and erase Claude’s memories directly. Users can also create ‘distinct memory spaces’ to keep different projects separate. This added functionality mirrors similar implementations in OpenAI’s ChatGPT and Google’s Gemini. Anthropic is also offering users the option to import memories from ChatGPT or Gemini, and all memories can be exported, avoiding vendor lock-in. However, some experts are raising concerns about the potential for chatbot memory to exacerbate existing issues, such as the risk of ‘AI psychosis’ and the amplification of biases due to the sycophantic nature of some AI models.