News · Published September 12, 2026

Claude Memory Now Works Across Chat and Cowork

Anthropic merged Claude's memory across its chat interface and Cowork, so something you mention in one now shows up when you switch to the other, instead of every session starting blank. TechCrunch and Engadget both reported the change on August 25, 2026, and it is enabled by default on Free, Pro, and Max plans across web, desktop, and mobile. For anyone using Claude for both quick questions and longer Cowork tasks, this closes a gap that used to mean repeating yourself.

What actually changed

Before this update, Claude's chat and Claude's Cowork mode kept separate memory, so a preference or project detail you mentioned in one didn't carry over to the other. Now the two share the same memory store. Mention a recurring project, a client's name, or a formatting preference in a normal chat, and Cowork can use it later without you re-explaining, and the reverse works too.

What it remembers, according to the reporting

TechCrunch's coverage describes the system capturing project details, stated preferences, and contextual information shared in conversation, things like a recurring headcount, a location, or a name you've mentioned before, and surfacing them again when relevant. It is not a transcript of everything you've ever said, it is closer to a running set of notes Claude keeps updating.

What stays out, by default

Per TechCrunch's report, Claude excludes sensitive categories from memory unless you explicitly opt in through an "include sensitive topics" toggle: health data, race, ethnicity, religion, political views, and gender identity. Some categories are excluded permanently regardless of settings, including government IDs, Social Security numbers, and criminal history. You can view and delete what's been remembered at any time.

How to use it

If you're on Free, Pro, or Max, the sharing is already on, nothing to enable. On mobile, TechCrunch notes you need an updated app to get it. If you want to check what's been saved, look in your account's memory or data settings, where entries are listed individually and can be deleted one at a time or all at once.

What this doesn't do

This is based on third-party reporting (TechCrunch, Engadget) from August 25, 2026, not a first-hand test of every edge case, and Anthropic can adjust the feature's behavior or rollout after publication. Shared memory means information can surface in a context you didn't expect, chat and Cowork now draw from the same notes, so it's worth checking what's stored if you use Claude for both personal and work-related conversations on the same account.

Questions people are asking about this update

What did Anthropic actually change with Claude memory?

Claude's memory now works across both the regular chat interface and Cowork, so something you mention in one shows up in the other, instead of each starting from zero. TechCrunch and Engadget both reported the change on August 25, 2026.

Is Claude memory on by default?

Yes. According to TechCrunch's August 25, 2026 report, it is enabled by default on Free, Pro, and Max plans, across web, desktop, and mobile.

Does Claude remember sensitive personal information?

Not by default. Health data, race, ethnicity, religion, political views, and gender identity are excluded unless you turn on an explicit include sensitive topics toggle. Government IDs, Social Security numbers, and criminal history are never stored, according to TechCrunch's reporting.

Can I see or delete what Claude remembers about me?

Yes. The memory is stored as editable entries you can view and delete yourself, per TechCrunch's coverage of the update.

Do I need to update my app to get this?

On mobile, yes. TechCrunch reported the feature requires an app update on iOS and Android. On web and desktop it applies automatically.

Keep reading

Disclosure: this page is an independent reader resource. Some links are Amazon Associates links, which means we may earn a commission from qualifying purchases at no extra cost to you.