An authorisation issue in Apple Intelligence, addressed with improved entitlement checks, could let a remote attacker access sensitive user data. The flaw is fixed in iOS 27.2 and iPadOS 27.2. Because it is an entitlement-check defect in the on-device model stack rather than a memory-safety bug, it is the class of failure that hardening work on AI features has to close, and it shipped inside the same mass patch round as the rest of the 27.2 release.