Incident response
What we do when member data may have been reached by somebody who should not have it, or when members cannot reach their work. This is our policy.
What counts
Anything that exposes, changes or destroys member data without the account's say, and anything that stops members reaching their work.
What we do
- 1
Stop it
Close the way in: end the sessions involved, withdraw keys and links, and take part of the service offline if that is what stops it.
- 2
Find out what was reached
Read the account histories and our own records to learn what was reached, when, and by whom.
- 3
Tell the people affected
We tell the owner of every account affected by email: what happened, what data was reached, what we have done, and what they should do.
- 4
Fix the cause
Fix what let it happen, and restore from the nightly copy anything that was changed or destroyed.
- 5
Write it down
Write down what happened and what we changed because of it.
What an account owner can do at once
The owner of an account can end any person's sessions from the Security screen, require two-step sign-in for everybody, and read the account's history of sign-ins and changes.
Tell us
If you think something is wrong, tell us. Write to us through the contact page. A weakness you found is welcome too: vulnerability disclosure says how.
The rest of the Trust Portal
- Trust PortalWhat is true today, in one list.
- How data is keptWhere documents sit, who can read them, and what goes to an AI model.
- Access controlRoles, two-step sign-in, sessions and the account history.
- Backups and their encryptionWhat the nightly copy holds and how it is locked.
- Vulnerability disclosureHow to report a weakness, and what is in scope.
- Outside servicesThe kinds of service that touch member data.
- Security questionnaireAsk for it, under a non-disclosure agreement.
- Security and privacyThe whole of what is stored, in one page.