The End of Constant Rebriefing
Anthropic has announced a significant change to Claude’s memory system. This Claude memory update merges the memory used by chat and Claude Cowork, creating a single, continuous assistant that remembers what it learned in one area even when you engage with it in another. The update eliminates the frustrating need to constantly rebrief the AI on details it already knows.
For users who have developed project ideas over time while chatting with Claude, the Claude memory update means they no longer have to start over when switching to Cowork to enact changes. The experience now feels like one seamless assistant rather than two separate products.
What the Claude Memory Update Changes
The new system improves how users move between research and action phases. Previously, if you were planning a conference agenda or drafting a manager update in chat, you would have to re-input all the details when moving to Cowork. Now, Cowork already understands the references from past conversations.
For example, if you discussed conference logistics in chat, Cowork would already know:
The headcount
The city
The speakers
Other relevant planning details
You no longer have to manually re-enter this information.
Smarter Memory Management with the Claude Memory Update
Real-Time Memory Updates
Anthropic has also improved how Claude builds its memory. Instead of summarizing conversations when they end, Claude now adds topics to its memory as you chat. This makes it easier to move between chat and Cowork because new memories are updated and shared more quickly, even when the chat is ongoing.
The Claude memory update is enabled by default on Free, Pro, and Max plans across web, desktop, and mobile. iOS and Android users need to update their app to the latest version to access the new capabilities.
Full Visibility and Control
One of the most user-friendly aspects of this Claude memory update is the transparency it offers. Anthropic is now exposing what Claude has retained in its memory, allowing users to:
Read stored information
Edit any topic
Delete information as needed
This level of control helps users maintain accurate and relevant memory while managing their privacy.
Privacy and Sensitive Information
Anthropic has implemented thoughtful privacy protections with this update. By default, Claude won’t store personal or sensitive information such as:
Health data
Race or ethnicity
Religious beliefs
Political views
Gender identity
However, users can opt in to enable these topics by toggling on “include sensitive topics in memory” if they prefer. The app also notifies users whenever a sensitive topic is saved.
What Claude Never Saves
Anthropic has set firm boundaries on what Claude won’t save under any circumstances:
Government-issued IDs
Social Security numbers
Criminal history
Immigration status
Anything violating the acceptable use policy
This ensures users can benefit from persistent memory without worrying about their most sensitive data being retained.
Why This Claude Memory Update Matters for Productivity
This update addresses one of the most annoying aspects of using AI agents. The constant need to rebrief AI on context wastes time and breaks workflow momentum. With the memory merge, users can:
Maintain continuity across projects
Switch between planning and execution seamlessly
Avoid repeating themselves
Build on past conversations without starting from scratch
The Claude memory update represents a step toward truly helpful AI that understands context and retains useful information without requiring constant repetition.
Getting Started
The new memory system is available now. To take advantage of the Claude memory update:
Update your Claude app if you’re on iOS or Android
The memory feature is automatically enabled on all plans
Review your stored memories through the settings to ensure accuracy
Consider whether you want to enable sensitive topics
This Claude memory update makes Claude a more capable and helpful assistant, whether you’re using it for research, planning, or taking action in Cowork.
