Provide the count of breached accounts on a domain
As part of the API, provide the count of breached accounts on a domain in a time window. I realise that for the domain search, users need to prove ownership of the domain before receiving the list of breached emails, which certainly makes sense. If the count of breached accounts on a domain isn't deemed too sensitive to disclose, this would be useful in third party risk monitoring applications which could then display "50 accounts with emails on your domain @domain.com have been breached in the previous 3 years" for example.
I’m declining this one simply for the reason you’ve already highlighted: it’s too sensitive. For example, you could restrict the range to the time of the Ashley Madison data breach, feed in the domain of a small company and start to draw some pretty sensitive conclusions. As it stands, domain owners can already derive this info so there’s way more risk than upside to this one.
I am not an API user.
How about providing a count per calendar year for domains which have breeches associated with over over 1000 distinct addresses?
My org's domain has at least 100,000 users, and I'd like to be able to reiterate the dangers of password reuse. Getting this information from the domain controller seems remote.