Always excluded
- Sensitive identification numbers (SSN, government ID numbers)
- Criminal history
- Immigration status
- Anything that violates Anthropic’s Acceptable Use Policy
Signature experience · synthetic specimen
The memory you use in chat is the same memory Claude Cowork uses. It lives as short topic files under Topics in Memory settings, and you can read, edit, or delete any of them. On Free, Pro, and Max it is on by default across web, desktop, and mobile. Sensitive topics stay off until you opt in. On Team and Enterprise, admins gate availability and it stays off until you enable it.
The desk below is a labeled synthetic specimen so you can practice that control without opening a real account. Independent educational publication. Not affiliated with Anthropic.
Two gates, not one
On Team and Enterprise, an admin decides whether memory exists for the organization. It still stays off for you until you turn it on. Click the gates in order.
Topics · synthetic
Five work chips on. Sensitive off. Managed chip waiting on a plan that is not this one.
Memory is off on this managed account. Enable the admin gate, then enable it for yourself, and the topic chips return to the blotter.
Arrow keys move, enter inspects, E edits, ⌫ deletes, S toggles sensitive topics when the desk has focus. Edits stay in this browser tab.
Always excluded
Sensitive, off by default
Health, race, ethnicity, religious beliefs, politics, gender identity, and similar areas. Opt-in saves going forward only.
Not this page
Context-window size, compaction, and prompt caching live onclaudecontext.com. Memory everywhere is not a larger window.
How to use the real list
Step 1
Go to Settings > Memory. That is the control surface for turning memory on, browsing topics, and choosing whether sensitive topics are included.
Step 2
Everything Claude remembers appears as short files under Topics. Skim them the way you would skim a small config directory.
Step 3
Correct the topic file rather than arguing in the next chat. A fix pays off everywhere, including Cowork cloud tasks.
Step 4
Remove topics tied to completed projects instead of resetting all memory. Reset is still available, and it is destructive.
Step 5
Health, beliefs, politics, and similar categories stay out of memory until you enable them. The opt-in is not retroactive, and some categories never save.
Step 6
Confirm an admin has made memory available, then enable it for yourself. It stays off until you do.
Anthropic’s Claude's memory works everywhere, and you decide what's in it is the primary source for this desk. The memory you use in chat is the same memory Claude Cowork uses. A preference explained once is available when Cowork runs a task in the cloud, and what comes up in that task carries back. That is why a stale topic is more expensive than it used to be: it now reaches both conversations and cloud work.
The stored unit is a topic file, not a rolling after-the-fact summary. Claude adds topics while you chat. Under Topics in Memory settings, each file is short, and each one can be read, edited, or deleted on its own. Correct a company name once and every later conversation, including Cowork, gets it right.
On Free, Pro, and Max, memory is on by default across web, desktop, and mobile. Saving sensitive topics remains off on those plans until you opt in. On iOS and Android you need the latest app version.
On Team and Enterprise, control moves up a level. Admins decide availability for the organization, and memory stays off for individual users until they turn it on. If you are on a managed plan and the topic list is empty, check the admin gate before assuming a bug. The desk’s Team / Enterprise mode is there so that sequence is visible.
By default Claude does not store topics related to health, race, ethnicity, religious beliefs, politics, gender identity, or similar areas. The official example of a useful opt-in is a gluten allergy when suggesting recipes. Turn on include sensitive topics in memory and Claude saves those topics going forward only. Nothing from before is backfilled, you get a notice each time something in those categories is saved, and you can turn the setting off again.
Some categories stay out even with the setting on: sensitive identification numbers, criminal history, immigration status, and anything that violates the Acceptable Use Policy. Those never become chips. The dietary-notes chip on this desk is a synthetic stand-in for the opt-in path, labeled as such, and is not a real medical record.
Memory everywhere is not a larger context window and not a caching feature. It changes what Claude knows at the start of a conversation or task, not how much it can hold during one and not what a repeated prompt costs. For windows, compaction, and prompt caching, useClaude Context.
It is also not Claude Code memory. CLAUDE.md andauto memory are files on your machine and in your repository. They are not the chat and Cowork topic list, and this announcement does not change them.
FAQ
No. They are a labeled synthetic specimen authored by Claude Memories. Every topic name and file body is invented. The page does not connect to a Claude account, does not read Settings > Memory, and does not upload anything.
Yes. Anthropic’s September 2026 announcement states that the memory you use in chat is the same as in Claude Cowork. Context built in conversation is available when Cowork runs a cloud task, and what comes up in Cowork carries back to chat.
Yes. Memories are stored as individual topic files under Topics in Memory settings. Each one can be read, edited, or deleted on its own. Pausing or resetting all memory remains available.
Sensitive topics such as health, race, ethnicity, religious beliefs, politics, and gender identity are excluded by default. You can turn on “include sensitive topics in memory” in Memory settings. That opt-in saves going forward only; nothing from before is backfilled. Some categories stay excluded either way.
On Team and Enterprise, admins control availability for the organization, and memory stays off for individual users until they turn it on. That is two gates, not one. On Free, Pro, and Max, memory is on by default across web, desktop, and mobile.
No. It changes what Claude knows at the start of a conversation or Cowork task, not how much it can hold during one and not what a repeated prompt costs. Context-window sizing, compaction, and prompt caching belong on claudecontext.com.
No. Claude Memories is an independent educational publication and is not affiliated with Anthropic, Claude, or Claude Code.
Cite this page
Claude Memories. "The Chips: inspect, edit, and delete Claude memory topics." claudememories.com, updated 2026-09-19. https://claudememories.com/chips/
The topic chips are original work by Claude Memories and are labeled synthetic. Cite them as an illustrative artifact, never as observed contents of a real Claude account. Claude Memories is an independent educational publication and is not affiliated with Anthropic.