JWT Attack Techniques - Security Vulnerabilities & Exploitation

Exploit servers that accept tokens with the "none" algorithm, allowing unsigned token modification.

Convert an RS256 token to HS256 and sign it with the public key as a secret.

Manipulate the kid header to point to malicious keys or exploit path traversal.

Redirect JWKS key fetching to an attacker-controlled endpoint.

Inject malicious claims to escalate privileges or bypass authorization.

Brute force weak or default secrets to forge tokens.

Reuse old tokens that should have been invalidated.

Take over a subdomain used for JWKS and serve malicious keys.

Use timing differences in signature verification to extract secrets.

Manipulate the x5u header to point to malicious certificates.

Extract JWT secrets from public repositories, documentation, or error messages.

Tokens passed in URL parameters can be logged, cached, or leaked via Referer headers.

Tokens exposed in error messages, logs, or debug output can be exploited.

Predictable jti values allow token replay or collision attacks.

Weak random number generation makes tokens predictable and vulnerable.

Some implementations may accept algorithm names with different cases, bypassing validation.

Inject nested JSON objects in claims to exploit parsing vulnerabilities.

Improper token caching can lead to stale tokens being accepted or valid tokens being rejected.

Failure to validate all claims allows injection of malicious or unexpected values.

Split tokens across multiple requests or use multiple tokens to confuse validation logic.

Exploit padding validation in encrypted JWTs (JWE) to decrypt tokens.