Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Federation lets one identity provider authenticate users to many services. That convenience concentrates risk: compromise the issuer, or its trust configuration, and everything downstream follows.

What Are Identity Provider Federation Risks?
Federation lets one identity provider authenticate a user once and assert that identity to many downstream applications, using SAML or OpenID Connect. The appeal is obvious: single sign-on, central policy, one place to disable a leaver. The consequence is equally obvious once stated. The identity provider becomes the control point for everything, and anything that compromises it, or the trust configuration around it, compromises the lot.
A service provider accepts an assertion because it is signed by a key it trusts. Obtain that key and you can mint valid assertions for any user, with any attributes, without touching a password or a second factor. On-premises AD FS makes this concrete—the technique called Golden SAML, seen in the SolarWinds intrusion disclosed in December 2020, involved theft of the token-signing certificate and the forging of assertions accepted as genuine by cloud services.
No failed login appears, no impossible travel alert fires, and conditional access policies evaluated at the identity provider are never consulted. Store signing material in a hardware security module, rotate it on a schedule and immediately after any suspected compromise of the hosting server, and monitor for assertions with unusual lifetimes, issuers, or attribute sets.
The second class of risk needs no cryptographic theft at all. An attacker with sufficient rights in the directory adds a new federated domain, registers an additional trusted issuer, or alters the attribute mapping so that a claim they control determines which account is matched. The backdoor then looks like ordinary configuration, and it survives password resets, key rotation, and most incident response.
Review federation trusts, issuer certificates, and claim transformation rules as deliberately as you review administrative group membership. Alert on any change to them. In Microsoft Entra ID, pay particular attention to cross-tenant access settings and to the creation of new federated domains; in AWS, to identity provider objects and the trust policies of roles that accept them.
Access tokens are bearer credentials. Once issued, they work for whoever holds them until they expire, and refresh tokens extend that for days or weeks. Adversary-in-the-middle phishing kits now routinely capture the post-authentication session rather than the password, which defeats most second factors entirely—the user authenticated correctly, and the attacker simply took the result.
Mitigations exist but must be switched on: phishing-resistant authentication with FIDO2 keys or passkeys, token binding and continuous access evaluation so a revoked session is actually terminated, short refresh token lifetimes for privileged roles, and conditional access that requires a compliant device for administrative actions. Our entry on multi-factor authentication covers the distinction between factors that resist phishing and factors that merely add friction.
Using a managed identity provider makes it a critical supplier. Okta’s disclosure in October 2023 of unauthorised access to files in its customer support system, affecting session tokens from customer-submitted diagnostic archives, is a reminder that the risk includes the vendor’s own operations as well as its product. Apply supplier assurance proportionate to the dependency: ask about their incident history, their support access model, and their notification commitments, and include them in the resilience planning expected under the NIS Regulations or DORA if those apply to you.
Keep a break-glass path that does not depend on federation—a cloud-native administrative account with hardware multi-factor authentication, excluded from conditional access policies that could lock it out, alerting on every use, and tested quarterly. If the identity provider is unavailable or untrusted, that account is how you regain control, as set out in cloud incident response readiness.
At Privia Security, we test federation trust and the escalation paths around it through our cloud security services.
On balance it improves security, because it centralises policy, removes password reuse across dozens of applications, and makes leaver removal reliable. What it does is concentrate risk: the identity provider becomes the one component whose compromise affects everything. That argues for treating it as a tier-zero asset with hardened hosting, rigorous change review, and strong monitoring, not for abandoning federation.
Three common ways. They steal the signing key and forge assertions, so authentication never happens. They capture the post-authentication session token with an adversary-in-the-middle phishing kit, so the second factor was satisfied by the real user. Or they use a flow that does not prompt, such as device code phishing or an OAuth consent grant. Phishing-resistant factors and token protection address the second and third.
Changes to federation trusts, issuer certificates, federated domains, and claim mapping rules; new or modified service principals and their credentials; consent grants with broad scopes; administrative role assignments; use of the break-glass account; and authentication patterns that bypass conditional access. Those changes are rare in normal operation, which makes them high-signal alerts rather than noise.
Yes. A cloud-native break-glass account that does not depend on the identity provider is how you recover when federation is broken or compromised. Protect it with hardware multi-factor authentication, split the credentials under dual control, exclude it from conditional access policies that could lock it out, alert on every use, and test it on a schedule so you know it still works.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.