AI0-001 AI Security Practice Question
A security analyst discovers that an attacker has been querying a production LLM API with thousands of carefully crafted prompts and using the responses to build a local copy of the model. Which attack is occurring?
⚠ Common exam trap
AI0-001 often tests the distinction between inference-time attacks (extraction, membership inference, evasion) and training-time attacks (poisoning); candidates confuse model extraction with prompt injection because both involve querying the API.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Model extraction
Model extraction (also called model stealing) occurs when an attacker queries a deployed model API repeatedly with crafted inputs and uses the input-output pairs to train a surrogate model that approximates the target's behavior. The scenario—thousands of queries used to build a local copy—is the textbook definition. The goal is to replicate the model's functionality, often to avoid API costs or to enable further attacks like adversarial example transfer.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Prompt injection
Why it's wrong here
Prompt injection manipulates a model's behaviour through crafted input; it does not extract training data or model weights. It is tempting because both involve adversarial prompts, and prompt injection would be correct where an attacker overrides system instructions or exfiltrates another user's context.
- ✓
Model extraction
Why this is correct
Model extraction directly fits: the attacker abuses legitimate API access, harvesting input-output pairs at scale to train a substitute model that replicates the victim's behaviour. This satisfies the stem's constraint of thousands of crafted queries whose responses build a local copy, distinguishing it from prompt injection or jailbreaking, which target content rather than model replication.
- ✗
Data poisoning
Why it's wrong here
Data poisoning corrupts the training pipeline by inserting malicious samples before or during training, not querying a deployed API. It is tempting because both attacks target model integrity, and poisoning would be correct where an attacker controls or contaminates the fine-tuning dataset.
- ✗
Membership inference
Why it's wrong here
Membership inference determines whether a specific record was in the training set, so it cannot describe an attacker reconstructing the model's behaviour from API outputs. It is tempting because it also targets model confidentiality, and would be correct if the attacker sought to confirm an individual's data was used in training.
About these practice questions
Courseiva writes every AI0-001 question from scratch — 962 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official CompTIA exam blueprint
This AI0-001 practice question is part of Courseiva's free CompTIA certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI0-001 exam.