Skip to main content

Claude's Memory Now Syncs Across Cowork and Chat

Anthropic just rolled out a significant update to Claude's memory system, and it's all about making your AI interactions feel less like starting from scratch every time. The company has integrated the memory functions of its Cowork and chat features, meaning the AI can now remember details from one and apply them to the other. No more repeating your project background or preferences every time you switch between the two.

What's Actually Happening?

Think of it this way: before, if you told Claude in chat that you're working on a marketing campaign for a new coffee brand, you'd have to remind it again when you moved to Cowork to draft a report. Now, that context carries over automatically. The official announcement highlights that Cowork now has memory, and it can reuse the same memory in chats. So, whether you're brainstorming in chat or executing tasks in Cowork, Claude has a continuous thread of your ongoing work.

This bidirectional integration works like a two-way street. When Cowork is handling tasks in the cloud, it can pull up information you've shared in chats. And any new memories Cowork generates during its work can be referenced later in your conversations. It's a seamless loop that aims to make your AI assistant feel more like a collaborator who remembers your history, not a stranger you meet for the first time every day.

Privacy First: Sensitive Stuff Stays Off by Default

Now, you might be wondering: "Does this mean Claude is remembering everything about me?" Not exactly. Anthropic has been careful to keep sensitive topics—like health details, personal beliefs, or other private matters—out of the default memory. These are disabled unless you manually opt in through the "Memory" settings. So, if you want Claude to remember your dietary restrictions or your stance on a particular political issue, you'll need to flip that switch yourself.

The memory feature is enabled by default for Free, Pro, and Max plans across web, desktop, and mobile. But remember, sensitive items stay off unless you turn them on. For iOS and Android users, you'll need to update to the latest version to see these changes.

What This Means for You

For heavy users, this update is a game-changer. It turns your "personal context" into a persistent asset that follows you across different workstations. You can start a conversation on your phone, continue it on your desktop, and pick up right where you left off in Cowork—without having to re-explain everything. It's like having an assistant who actually remembers your preferences and past projects.

But it also puts the control back in your hands. You decide what Claude remembers long-term. The default is to keep things safe, but you have the flexibility to expand its memory if you trust it with more personal details.

The Bigger Picture

This move by Anthropic signals a broader trend in AI development: making assistants more context-aware and personalized. The goal is to reduce friction in human-AI interaction. Instead of treating each session as a blank slate, AI is learning to build on previous interactions, creating a more natural and efficient workflow.

Of course, this raises questions about privacy and data security. Anthropic's approach—keeping sensitive info off by default—shows they're aware of these concerns. But as AI memory becomes more integrated, users will need to stay vigilant about what they share and how they manage their settings.

Key Points

  • Unified Memory: Claude's Cowork and chat now share a single memory system, allowing context to flow between them.
  • Bidirectional Integration: Memories from chats inform Cowork tasks, and vice versa.
  • Privacy Defaults: Sensitive topics (health, beliefs) are not stored by default; users can enable them manually.
  • Availability: Enabled by default for Free, Pro, and Max plans on web, desktop, and mobile; iOS/Android users need the latest update.
  • User Control: You decide what Claude remembers long-term, balancing convenience with privacy.

Image