Daily AI CatchupAnthropic merges Claude and Cowork memory, now synced by default across platforms
Anthropic unified Claude's memory system across its chat and Cowork agent platforms, so conversations and context persist regardless of which interface users access. The feature is on by default. Claude now proactively adds topics to memory during conversations, and everything is stored as a readable/editable list of files under Topics in memory settings. Users retain full control to read, edit, or delete memory entries individually.
Why it matters
💻 Developer · Unified memory architecture simplifies building persistent agent experiences. You no longer need to manage separate conversation stores for chat vs. agents—memory becomes a single coherent layer. This is foundational for long-horizon workflows where context needs to carry across interaction modes.
📦 Product · Continuity across chat and agents increases perceived intelligence and reduces friction. Users don't lose work switching between interfaces. The opt-out mechanism (readable/editable memory) also addresses privacy concerns without sacrificing the feature—a smart default-on approach.
🎨 Design · Users see their AI truly learning their work without magic. Making memory visible and editable builds confidence and transparency. The proactive topic-tagging during conversations is thoughtful: memory surfaces without requiring explicit commands.
📈 Business · Stickier users. The more Claude remembers about their work, the harder it is to switch. This is a retention lever that's harder for competitors to copy if they don't have equivalent agentic infrastructure. Memory becomes a moat.
🤔 Just Curious · This signals Anthropic's long-term bet on agents. By making memory central and visible, they're treating it as a first-class feature, not a hack. It's a statement that persistent, multi-modal interaction is core to how AI will be used.
Sources: Anthropic merges Claude chat and Cowork memory, on by default