Which type of attack attempts to find data by repeatedly trying similar SQL with a modified predicate?

Question

Options

• Atiming attack
• BInference attack
• Cdata remanence attack
• Dcache attack
• Eside-channel attack
• Fknown-plaintext attack

• A
  4% (2)
• B
  2% (1)
• C
  8% (4)
• D
  72% (38)
• E
  2% (1)
• F
  13% (7)

Explanation

There appears to be an error in the provided answer key - the correct answer for this question should be B. Inference attack, not D.

Why B is correct: An inference attack against a database works by submitting repeated SQL queries with modified predicates (e.g., WHERE salary > 50000, then > 60000, etc.) to deduce sensitive data without directly accessing it. The attacker observes responses - rows returned, error messages, true/false behavior - and narrows down the actual values. This is the textbook definition of inferential (including blind) SQL injection.

Why the distractors are wrong:

• A. Timing attack - exploits differences in execution time to infer secrets, not SQL predicate manipulation specifically.
• D. Cache attack - exploits CPU/memory cache behavior (e.g., Flush+Reload) to leak data; has nothing to do with SQL predicates.
• E. Side-channel attack - the broader category that includes timing and cache attacks; not predicate-based SQL probing.
• C. Data remanence - data persisting after deletion (e.g., unwiped disk sectors); unrelated.
• F. Known-plaintext - a cryptographic attack where the attacker has both plaintext and ciphertext; not a database query technique.

Memory tip: "Inference = I guess by asking yes/no questions." If an attack builds knowledge by repeatedly tweaking a WHERE clause and watching what changes, that's inference. Cache and timing attacks live at the hardware layer, not the SQL layer.

Note for exam prep: If this comes from an official practice exam, I'd flag it - the answer key appears incorrect. Check your course material for how "inference attack" is defined in that specific context.

No community discussion yet for this question.