This is a **practice assessment**. It is designed to sharpen judgement and create CPD-friendly reflections. It is not a timed exam.
AI practice assessment. Modern systems
12 questions
Scenario: An agent can call tools. What control limits blast radius fastest?
agents
Why must LLM systems log tool use?
monitoring
Governance is most defensibly framed as:
governance
Scenario: You must store evidence but protect privacy. What is the defensible approach?
evidence
A system passes a benchmark but fails in production. What was tested incorrectly?
monitoring
Scenario: An agent starts looping tool calls. What is the safest immediate control?
agents
What makes an AI incident response plan credible?
governance
Which signal is most useful for detecting retrieval problems early?
monitoring
Scenario: Leadership wants 'one model for everything'. What is the first system-level risk to raise?
risk
Scenario: You need to prove safe operation to an auditor. What evidence is most defensible?
operations
Scenario: Product wants to disable refusal behaviour because it 'hurts conversion'. What is the governance response?
governance
Scenario: An agent suggests a plan. What is the most defensible way to execute it?
agents
Add CPD reflection (optional)
One short paragraph makes your CPD evidence much stronger.
If any answer surprised you, write a one-paragraph note: what assumption changed, what evidence you would look for, and what you would do differently next time.
Quick feedback
Optional. This helps improve accuracy and usefulness. No accounts required.
