Changing behavior based on user agent is necessarily intentional on the part of Microsoft. That check lies somewhere along the line between "having the direct

Changing behavior based on user agent is necessarily intentional on the part of Microsoft. That check lies somewhere along the line between "having the direct goal of breaking authentication flow (pure malice)" and "is a completely legitimate programming error (pure incompetence)." I am not ready to assume pure incompetence (and here's where I might be wrong).