Anthropic has merged Claude chat and Cowork memory.
Claude now remembers what you shared in one area and uses it in another.
Users can view, edit, or delete stored information.
What happened
Anthropic announced on Tuesday that it is merging the memory system used by chat and Claude Cowork, so Claude will remember what it learned in one area when used in another. Users can also read, edit, or delete information Claude has retained.
Why it matters
Previously, users had to rebrief Cowork with details from past chats, which was frustrating. Now Claude feels like one continuous assistant, making it easier to move between research and action without re-entering information like headcount, city, or speakers.
What to watch
The memory feature is on by default for Free, Pro, and Max plans across web, desktop, and mobile. On iOS and Android, users must update the app to the latest version. By default, Claude won't store sensitive personal data unless users opt in.
Ask the AI about this article →
The update addresses a common pain point: the need to repeatedly re-explain context when switching between research and action. By sharing memory between chat and Cowork, Anthropic aims to make Claude behave as a single, continuous assistant, reducing friction for users who develop plans in conversation and then execute them elsewhere.
This change also gives users more control over their data, with the ability to review and delete what Claude has stored. The default exclusion of sensitive categories like health, politics, and immigration status suggests an effort to balance personalization with privacy, while still allowing users to opt in if they prefer.
While the announcement focuses on usability, it may also signal Anthropic's push to integrate its AI more deeply into everyday workflows, particularly in business settings where project details and context are often scattered across multiple interactions.
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytime
Ask AI anything about this article. Q&As are published on this page for other readers too.
Anthropic PBC announced it is changing how Claude, its flagship AI product, uses memory, allowing users to see…
A design principle for AI agents is proposed: daily assistants should have a 24-hour session that resets at mi…
Google is launching Gemini Enterprise for Legal, a product that uses MCP connectors to link Google's AI model…
More than one million people clicked LinkedIn's 'seems like AI slop' button in the first two weeks of its laun…
A guide explains six job-hunting steps that AI agents can handle, from self-analysis and salary research to re…
Amazon OpenSearch Service now supports MCP Apps, which extend the Model Context Protocol so each tool call ret…
