Cybersecurity glossary

What is a Claim?

Learn what a claim is in identity and access security, how assertions convey attributes about a subject, where claims appear in JWT SAML and OIDC, and how to validate them safely.

Identity and accessUpdated July 23, 2026
Also known asIdentity claimSecurity claimAttribute assertion

Definition

In identity and access security, a claim is an assertion made by an identity provider or authorization server about a subject—such as an identifier, role, group membership, or authentication strength—that relying parties may use after validating the issuer and integrity of the assertion.

Why claims matter

Identity systems need a portable way to say facts about a principal. A claim is that fact: an issuer-backed assertion that a relying party can evaluate after it trusts the assertion’s authenticity.

Claims power single sign-on, API authorization, and zero-trust decisions. They also create risk when applications confuse “data present in a request” with “data attested by a trusted issuer.”

What a claim represents

Subject identity

Stable identifiers that say who the assertion is about.

Context attributes

Email, department, tenant, device posture, or assurance level.

Authentication properties

How and when the subject authenticated, including MFA evidence.

Authorization hints

Roles, groups, and scopes that policies may consume.

From assertion to authorization decision

1

Issuer authenticates the subject

An IdP or authorization server verifies credentials or federation.

2

Claims are assembled

Directory attributes and session context become assertion fields.

3

Assertion is protected

Signatures, encryption, or secure channels protect integrity and sometimes confidentiality.

4

Relying party validates the issuer

Audience, signature, certificates, and issuer identity are checked.

5

Claims are normalized

Names and value formats are mapped into the app’s policy model.

6

Policy decides access

Local rules combine claims with resource and action context.

Claims across common protocols

ProtocolClaim carrierExample fields
JWT / OAuthAccess token payloadsub, scope, aud, exp
OpenID ConnectID token / UserInfosub, email, acr, amr
SAMLAssertion AttributeStatementNameID, group attributes
Session systemsServer-side session recorduser id, roles, auth time

Safe claim handling checklist

  • Trust claims only from configured issuers after integrity validation.
  • Separate authentication claims from authorization policy evaluation.
  • Prefer short lifetimes for high-churn attributes like roles and groups.
  • Reject client-supplied security attributes in request bodies.
  • Document claim catalogs and naming conventions for every application.
  • Watch for over-exposure of PII inside portable assertions.
  • Re-evaluate critical entitlements when assurance requirements change.
  • Test how apps behave when mandatory claims are missing or malformed.

The practical takeaway

A claim is an issuer’s statement about a subject. Protocols differ, but the security rule is constant: validate who asserted it, then decide what it means for access.

For JWT-specific encoding and registered fields, continue with JWT Claim and the individual aud, exp, iat, iss, nbf, and sub pages.

Related security terms

Frequently asked questions

What is a claim in simple terms?

A claim is a statement about a user or system—like “this is Alice,” “she is in Finance,” or “she signed in with MFA.” Apps trust those statements only after checking who issued them and that they were not altered.

Are claims the same as permissions?

Not exactly. Claims are assertions of attributes. Permissions are authorization decisions that may be derived from claims plus local policy.

Where do claims appear?

Commonly in JWT payloads, OpenID ID tokens, UserInfo responses, SAML assertions, and proprietary session tokens.

Who is allowed to make claims?

Only issuers your application is configured to trust. A claim from an untrusted party is just unauthenticated data.

What is claim-based authorization?

An approach where access rules evaluate attributes from validated tokens—roles, groups, tenant IDs—rather than only local account lookups.

Can claims be stale?

Yes. Embedded role or group claims can lag behind directory changes until the token expires or is refreshed.

Should apps accept claims from the client body?

No. Security claims must come from validated issuer assertions, not from request JSON the client can freely edit.

References

Explore authoritative guidance and frameworks related to claim.

Explore every security definition

Return to the glossary to search by term, alias, starting letter, or security category.

Browse glossary