These perspectives are not opposing. Sounds very much like they have AI or unqualified humans manking final decisions on abuse reports/scans and then refuse to reevaluate when the customer complains. So just like most tech companies.