AISG practice

25 questions. Choose the BEST answer.

You can answer as a guest, but you'll need a free account to see your score and save it. Sign up · Sign in
Question 1 Question 1 · M2
Which statement BEST distinguishes transparency from explainability?
Question 2 Question 2 · M2
A declined loan applicant asks why. Which explanation type is MOST useful to them?
Question 3 Question 3 · M3
A widely used open model is found to contain a backdoor. Which artefact MOST quickly tells you which of your systems are affected?
Question 4 Question 4 · M8
Which SOAR action is an appropriate automated first response to a high-confidence exfiltration pattern from an agent session?
Question 5 Question 5 · M5
Why should production personal data be kept out of AI development environments?
Question 6 Question 6 · M5
In a canary rollout, what should happen if safety-filter hits spike?
Question 7 Question 7 · M7
Text in white-on-white inside a PDF instructs an assistant to leak data. Which threat category is this?
Question 8 Question 8 · M8
How should a new jailbreak technique be tracked?
Question 9 Question 9 · M5
Which model-card section should state that a model must not be used for employment decisions?
Question 10 Question 10 · M7
Which attack reconstructs representative training data, such as a face for a class label?
Question 11 Question 11 · M5
A hosted model provider will retire your current model version next month. What is the BEST preparation?
Question 12 Question 12 · M5
Why must evaluation jobs and test sets be protected from the training pipeline's write access?
Question 13 Question 13 · M6
Streaming model output token by token to users primarily complicates which control?
Question 14 Question 14 · M7
Which mitigation for an injection-driven payee-change finding is BEST?
Question 15 Question 15 · M7
Which knowledge base catalogues adversary tactics and techniques against AI systems, modelled on ATT&CK?
Question 16 Question 16 · M3
A vendor's SOC 2 Type II report pre-dates the launch of the AI feature you plan to buy. What is the BEST next step?
Question 17 Question 17 · M5
Which API design change MOST reduces the information available for extraction?
Question 18 Question 18 · M1
An email assistant agent runs with a service account that can read every mailbox. A user asks it to summarise 'my emails' and it retrieves another user's mail. This is an example of:
Question 19 Question 19 · M7
Which MAESTRO layer covers tool misuse and unsafe plugins?
Question 20 Question 20 · M6
What is context boundary validation?
Question 21 Question 21 · M2
An Indian bank uses a vendor's credit-scoring model for EU customers without modifying it. What is the bank's role under the EU AI Act?
Question 22 Question 22 · M7
Which defence MOST directly reduces membership inference risk?
Question 23 Question 23 · M3
Mirroring approved models into an internal registry PRIMARILY helps to:
Question 24 Question 24 · M2
An operations team wants an agent that can place purchase orders autonomously up to a set limit. Based on the intake answers, which route is MOST appropriate?
Question 25 Question 25 · M1
A team plans to run inference on regulated health data using shared cloud GPUs. Which control MOST directly addresses residual data exposure between tenants?