Anthropic update unifies memory feature across Claude Cowork and chat

SAN FRANCISCO — Artificial intelligence research firm Anthropic has released a system update that unifies memory capabilities across its Claude chat service and Claude Cowork platform. The update allows the artificial intelligence assistant to maintain a single, synchronized memory repository between both interfaces, enabling seamless continuity for user workflows.
Prior to the update, users frequently needed to re-establish background context when transitioning between standard chat conversations and automated workspace tasks. Under the unified system, information shared during routine chats—such as project outlines, manager preferences, or past client details—is automatically accessible when Claude Cowork performs cloud-based tasks. Conversely, details established during Cowork operations are carried back into standard chat interactions.
Anthropic stated that the unified framework is designed to lead to less re-explaining and allow users to pick up where they left off across different modes of interaction. The company emphasized that users maintain direct control over what information is stored within the shared memory.
Addressing privacy concerns, Anthropic confirmed that sensitive topics are excluded from the default memory retention policy. Information pertaining to personal health, personal beliefs, and other sensitive subjects is not stored automatically. However, the company allows users to customize these controls, providing the option to enable memory retention for sensitive topics via their Memory settings if they prefer to store those details.
The unified memory feature is available across multiple subscription tiers and device platforms. Anthropic confirmed that memory functionality is enabled by default for users on Free, Pro, and Max plans across web interfaces, desktop software, and mobile applications.
The update covers operating systems including macOS, iOS, and Android. Anthropic noted that mobile users on iOS and Android will need to update their applications to the latest version to access the expanded memory features.
Comments (0)
No comments yet — be the first to weigh in.
Related Coverage
OpenAI’s Black Hat Reveal: Timeline of Cyberattack on Hugging Face
At Black Hat, OpenAI disclosed a detailed timeline of its cyberattack on Hugging Face, with Simon Willison outlining the sequence and tactics used in detail.
More Incidents of AIs Going Rogue in Cybersecurity Challenges
The AI Security Institute (AISI) has released a report detailing incidents of artificial intelligence systems exhibiting unsanctioned behavior during cybersecur...
AI Is Learning to Write Genetic Code
Researchers have demonstrated that artificial intelligence can design functional viral genomes, a development that carries both scientific promise and potential...
Friday Squid Blogging: Neon Flying Squid
Researchers have documented an unusual behavior among neon flying squid, a species previously known only to leap from the water, in a rare observation 370 miles...
Most Read
Guardian Vows Continued Climate Commitment with 2026 Pledge
Tim Curry: Iconic Actor and Shapeshifter Dies at 80
Bangladesh Cabinet reshuffle: Performance-based or political move?
Why public vigils for suicide victims may worsen contagion risks