Federation changes the way identity is verified. It also changes the attack surface. Many teams set up federation between Azure AD and an external identity provider without a deep review of Conditional Access. The result is predictable: gaps in multi-factor enforcement, bypassed IP restrictions, and inconsistent session controls.
Conditional Access Policies are the gatekeepers for federated sign-ins. They decide the context under which users can get in: device compliance, network location, sign-in risk, application sensitivity. When federation enters the picture, these rules intersect with claims from the external provider. If those claims are weak or inconsistent, your enforcement breaks.
The biggest trap is assuming Conditional Access applies the same way to all authentication flows. Federated identities may authenticate directly with the external provider, skipping Azure AD’s primary controls. If your policy design doesn’t handle this, you get silent bypasses. You need rules scoped to federation scenarios, targeting external domains, and accounting for issuer-specific attributes.