AISP practice · Module 1

AI Concepts and Security Fundamentals — 15 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 · M1
Which practice BEST reduces the risk of model collapse?
Question 2 Question 2 · M1
Which property of transformer models is the root cause of prompt injection?
Question 3 Question 3 · M1
An insider can edit the labels used to train a supervised malware classifier. Which outcome is MOST likely?
Question 4 Question 4 · M1
An investigator re-asks a chatbot the question that produced a harmful answer and gets a safe reply. What can be concluded?
Question 5 Question 5 · M1
A model's inputs look the same as at launch, but the relationship between inputs and the correct outcome has changed after a regulation change. What type of drift is this?
Question 6 Question 6 · M1
A system prompt says 'never reveal personal data', yet the model sometimes does. What principle explains this?
Question 7 Question 7 · M1
Why does overfitting create a privacy risk?
Question 8 Question 8 · M1
Why should an AI team log the exact retrieved chunks and prompt version for each response?
Question 9 Question 9 · M1
A team quantizes a production model to cut inference cost. Nothing else changes. What should happen before it replaces the original?
Question 10 Question 10 · M1
A fraud model is reported as 99.4% accurate on data where 0.5% of transactions are fraud. What should you ask for?
Question 11 Question 11 · M1
An HR chatbot uses RAG over the HR library and answers a contractor's question with a manager's salary from a spreadsheet in the same library. Which single control would have prevented this?
Question 12 Question 12 · M1
Reviewers override a loan model's recommendations in 0.1% of cases, against 6% during testing. What is the MOST likely concern?
Question 13 Question 13 · M1
An agent that reads customer emails can also issue refunds. What is the FIRST design step to reduce risk from injected instructions?
Question 14 Question 14 · M1
A team wants an assistant to answer questions about product terms that change monthly and differ by customer segment. Which approach fits BEST?
Question 15 Question 15 · M1
A security architect says a fraud model is 'only code', so the normal code review process is enough. What does this miss?