It would be reasonable for exchanges to parse Twitter feeds and other social channels for anti money laundering and fraud signal, similar to Github shipping AWS secrets accidentally exposed in commits to AWS for triage/suspension.
Once you’ve got the infra in place, you can have AML and other compliance staff triage and action from a dashboard (blocking suspect transfers until further review has been performed, and releasing transfers of a review shows nothing suspect).
(Have done some AML/KYC work in the fiat finance space)
One thing I've learned in life is that nothing is as polished or automated as you'd think. I would be surprised if anybody was doing this except maybe high tier law enforcement.
Only because I've seen first-hand how advanced their taint analysis is, so I'm already over that surprise.
Once you’ve got the infra in place, you can have AML and other compliance staff triage and action from a dashboard (blocking suspect transfers until further review has been performed, and releasing transfers of a review shows nothing suspect).
(Have done some AML/KYC work in the fiat finance space)