PT0-002 Attacks and Exploits Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✓

  JWT algorithm confusion attack (alg:none)

  Why this is correct

  In an alg:none attack, the tester modifies the JWT header to set the algorithm to 'none' and removes the signature from the token. If the server-side JWT library is misconfigured to accept the 'none' algorithm or fails to enforce an explicit algorithm allowlist, it will treat the token as valid without verifying any cryptographic signature. This effectively bypasses integrity checks and allows the attacker to forge arbitrary claims, making it a true algorithm confusion vulnerability.

• ✗

  JWT timing attack

  Why it's wrong here

  A JWT timing attack is a side-channel technique that measures response-time differences to infer a secret key or other sensitive data, such as during signature verification. It does not exploit the 'none' algorithm; rather, it requires the server to actually perform cryptographic verification and attempts to extract the secret through timing variations. Since the question involves bypassing signature verification via 'alg:none', a timing attack is not a relevant alternative explanation.

• ✗

  JWT kid injection

  Why it's wrong here

  In a 'kid' injection attack, the attacker manipulates the 'kid' (key ID) header field to point to an attacker-controlled key or a file path (e.g., SQL injection, path traversal), tricking the server into using that key for signature verification. This still requires the server to verify a signature using the substituted key; it does not eliminate the signature requirement. The 'kid' header controls which key is used, not the algorithm, so it cannot directly achieve an 'alg:none' bypass.

• ✗

  JWT brute-force of the secret

  Why it's wrong here

  Brute-forcing the JWT secret is an offline attack where the tester guesses the HMAC secret by trying many candidate values and comparing the resulting signatures, which is only relevant when the token is signed with a symmetric algorithm like HS256. If the server accepts 'alg:none', there is no signature to brute-force because the token is accepted without one. Thus, brute-forcing is unnecessary and does not explain the discovery of an unsigned token being accepted.

Go deeper

Related to this question