We sometimes put confidential information into task descriptions and elsewhere. Sending that off to an AI company is potentially bad. Doing it without an assurance that our data will not be added to their database for building responses is much worse.
I have turned the option off for my organization. It is great that you provide that option, but it should be off by default. Once it is a standard feature instead of a newly introduced feature (with email warnings), you will be potentially leaking confidential information for new customers unless there are big warnings at account creation and a few reminders after.
That is nice if true, but it is not specified anywhere I could find in the terms of service which is the governing document.
If you could point to where that AI privacy policy is specified (other than an unsigned mass email from no-reply@asnana.com) that would answer my suggestion.
You can follow the links in my post and that may get you what you’re looking for. One requires requesting access, as I mentioned, and I received a day or two later.