Practical steps to change what Claude remembers

Show summary Hide summary

Anthropic’s Claude organizes remembered information by topic, and users can edit, remove or limit what the assistant keeps. Knowing where to find those controls on the web and mobile helps you manage which details Claude stores and when it records more sensitive subjects.

Editing topic-based memories in the web app

  1. Open claude.ai and go to your profile menu, then choose Settings.
  2. Select Memory, then scroll to the Topics section.
  3. Use the on-screen icons to change a topic (Edit) or remove it (Delete).
  4. If you edit a memory, type instructions in the prompt bar to tell Claude what to change.
Laptop screen showing app settings menu to edit memory topics
Open the profile menu in the web app to access Memory settings.

Managing memories from your phone

On mobile, tap your profile icon to open Settings. Under the App heading pick Capabilities, then find and open Memory files.

Smartphone with app settings screen and profile icon visible
Tap your profile icon on mobile to find Memory files under Capabilities.

Choose a topic to view it. To change a stored item, describe the update in the prompt bar; to remove it, tap the trash icon.

How Claude treats sensitive information

By default, Claude does not save memories about certain sensitive areas such as health, race, ethnicity, religious beliefs, politics or gender identity. Anthropic recently added an opt-in labeled Include sensitive topics in memory. Turning it on permits Claude to start recording a wider set of personal details.

When Claude creates a memory that involves a sensitive subject, the interface will display a notice. Anthropic says older releases and some mobile builds may not show that prompt, and in those cases Claude will not record the sensitive topic.

Certain data remain excluded from memory under any setting. For example, Claude will never save government ID numbers, even if you enable sensitive-topic recording. The assistant also will not retroactively add items from past chats to its memory.

Stopping and clearing automatic memories

Claude generates memories automatically by default. You can stop new automatic entries by toggling off Generate memory from chats.

After disabling automatic generation, Anthropic offers two choices: Pause memory or Reset memory. Pausing halts further collection while keeping existing memories intact. Resetting wipes stored memories permanently — the action cannot be undone.

Anthropic notes that if you reactivate memory after a reset, Claude will start with a blank slate: when the feature is re-enabled, the assistant will not retain its previous memory.

Give your feedback

★★★★★

Be the first to rate this post
or leave a detailed review



The Pacific Tribune is an independent media. Support us by adding us to your Google News favorites:

Post a comment

Publish a comment