Anthropic has begun rolling out a memory search function for its Claude AI chatbot, giving users the option to retrieve and cite information from prior conversations. The capability is live for subscribers on the Max, Team and Enterprise tiers and will be extended to additional plans in the coming months, the company said on 11 August. The feature differs from rival systems by requiring an explicit prompt before Claude accesses historical exchanges, a design choice Anthropic says reduces the risk of unwanted data retention. Users can disable the setting entirely via a privacy toggle. Anthropic’s move follows OpenAI’s rollout in April of a broader ChatGPT memory tool that automatically draws on all past interactions, and Google’s tests allowing Gemini to personalize answers using search history. By limiting retrieval to targeted queries, Anthropic is positioning Claude as a more privacy-minded alternative while still narrowing the functional gap with competing chatbots.