In a move that will shock absolutely no one who has ever used an AI assistant, Anthropic has announced it's fixing one of the most irritating aspects of its Claude chatbot: the need to constantly re-explain things you've already told it. On Tuesday, the company revealed it's merging the memory systems of its chat and Claude Cowork features, meaning Claude will now remember what it learned in one place even when you're using it in another.

This update is a godsend for anyone who's ever spent hours chatting with Claude about a project, only to have to start from scratch when switching to Cowork to actually get things done. Previously, the two features operated as if they were estranged siblings who never spoke, forcing users to re-input crucial details like headcounts, city names, and conference speakers. Now, Claude will feel like one continuous assistant, rather than two products under one roof that refuse to acknowledge each other's existence.

Anthropic is also giving users a peek behind the curtain, letting them read, edit, or delete any information Claude has stored in its memory. No more wondering what the AI thinks it knows about you - now you can take a look and prune as needed. The company suggests this will make it easier to transition from research to action, like drafting an update for your manager or an agenda for a conference, without having to re-input all the context.

As if that weren't enough, Claude will now add topics to its memory as you chat, rather than waiting until the conversation ends to summarize. This means memories are updated and shared between chat and Cowork in real time, even if you're still mid-conversation.

But before you get excited about your AI overlord remembering your every whim, note that by default, Claude won't store personal or sensitive information like health data, race, ethnicity, religious beliefs, politics, or gender identity. If you want Claude to remember those, you'll have to toggle on "include sensitive topics in memory" - because who doesn't want their AI to know their political leanings? The app will also notify you whenever it saves a sensitive topic, so you can live in constant mild anxiety.

There are some things Claude will never remember, no matter how much you beg: government-issued IDs, Social Security numbers, criminal history, immigration status, or anything else that violates its acceptable use policy. So you can rest easy knowing your AI won't be filing your taxes anytime soon.

The memory feature is enabled by default on Free, Pro, and Max plans across web, desktop, and mobile. iOS and Android users will need to update their apps to the latest version, because of course there's an update.