Anthropic has introduced a memory feature to its Claude chatbot, allowing users to pick up conversations and projects without starting over. The tool lets Claude reference earlier chats on request, eliminating the need for users to repeat previous details. Anthropic also released a YouTube demo showing the feature in action.

In April, OpenAI made a similar move by enhancing ChatGPT’s memory capabilities, enabling it to recall past conversations and provide more personalized responses. In Anthropic’s demonstration, a user asked Claude what they had been working on before going on vacation. Claude retrieved past chats, summarized the topics, and asked if they wanted to resume the project.

“Never lose track of your work again,” Anthropic said, highlighting that Claude’s memory makes it possible to revisit old discussions, build on previous ideas, and manage projects without starting fresh. The feature works across web, desktop, and mobile, and can keep different projects or workspaces organized for quick access. It is currently rolling out to Max, Team, and Enterprise subscribers. To enable it, users can go to Settings → Profile → Search and reference chats. Unlike ChatGPT’s persistent memory, Claude does not automatically store all interactions or create a user profile. It only accesses past conversations when prompted.

In related news, Anthropic’s Claude Code now includes real-time security scanning for AI-generated code, detecting vulnerabilities like injection attacks and recommending fixes. Developers can enable an always-on mode to continuously monitor code changes, flag risks such as insecure data handling, and suggest improvements to prevent unsafe deployments.