Pausing AI recall is not the same as stopping capture
Understand the difference between collecting chats, using memory in answers, and making an explicit memory request in Dray.
Choose what gets shared 2 min read
Deleting AI memory vs. revoking access
Deletion concerns stored content. Revocation concerns who can access a service. Revoking a client does not by itself erase stored records, and deleting a record does not necessarily stop new capture.
Map the copies before choosing the action
A conversation can exist with the original provider, in a local archive, in a memory service, and in a destination chat after a handoff. Removing one copy is not evidence that all the others disappeared.
Write down which copy you want removed and which connections should stop working. This makes a support request more precise and helps you avoid treating a successful sign-out as a completed data deletion.
Check the result rather than promising completeness
After using an available removal control, check the affected view and record any confirmation. If you need to understand backups, retention periods, or downstream provider copies, ask for the active policy. Do not infer those details from an empty search result.
For material already sent to another AI provider, review that provider’s controls separately. A local action cannot guarantee removal of a response or copy held by a different service. Keep sensitive details out of the initial support message; describe the scope first.
TAKE THIS WITH YOU
A precise data request
An illustrative template. Replace the details with material you are allowed to share.
I want to remove [specified conversation or account data] from [service or storage location]. I also want [named connection] to lose future access. Please explain which copies this covers, any retention exceptions, and how completion is confirmed. Do not treat this as a request to remove unrelated projects.
USING THIS WITH DRAY
Keep the context. Check the source.
Use the current Dray data notes and available controls for removal or revocation. Contact [email protected] for data requests and questions about the active retention policy.
Dray is in beta. Available sources and connections depend on your setup. Read the data notes before granting access.
Sources & next reads
Product details are grounded in the guides below. Example projects and prompts are illustrative, not customer results.