Claude Cowork finally remembers what you told the app in chat
26 August 2026 · filed under 4b53c711439f
Anthropic has updated Claude so that context shared in chat now carries over into Cowork, according to TechCrunch AI. The company’s stated aim is to end the practice of users re-briefing the assistant each time they switch modes, a friction point that has long shaped how people work with conversational AI tools.
Per the report, Claude will draw on a shared memory layer spanning both interfaces, so that project details, preferences, and other context established in one setting persist into the other. TechCrunch frames this as a practical step rather than a dramatic leap, addressing a specific complaint rather than introducing a new capability category.
The report characterizes persistent memory as a foundational feature for AI systems generally, one that many assistants have lacked in consistent form across different modes of use. Anthropic’s move brings Claude’s chat and Cowork products into closer alignment on this point, reducing the repetition previously required of users moving between them.
No additional detail on rollout timing, technical architecture, or user controls accompanies the report as summarized. The change is described strictly in terms of behavior, information given to Claude in one context becoming available in another, without further elaboration on the underlying mechanism.
The Specola notes this item as a documented product update reported by TechCrunch AI, based solely on the material provided, with no independent confirmation beyond that single source.