AI Governance Assessment: Sample Questions | Bryq

Updated:

Ethical AI

AI Best Practices

AI Governance Assessment: Sample Questions | Bryq

Updated:

Ethical AI

AI Best Practices

These are two sample questions from Bryq’s AI Governance assessment, a test for hiring people who use, manage or oversee AI responsibly at work. The full assessment covers two skills: ethical AI and AI best practices. Each question describes a realistic situation involving AI systems or tools and asks which response is most appropriate, with four options to choose from.

Ethical AI

Intermediate

Question

Your team plans to use an AI model to decide which customer complaints are handled first. In testing, you find it consistently ranks complaints written in non-native English lower, regardless of how serious they are. What should you do?

Select your answer

Pause the rollout, find out why the model penalizes these complaints, and fix it before deployment.

Launch the model, and monitor complaint outcomes for signs of unfair treatment.

Launch the model as planned, since it ranks most complaints accurately.

Launch the model, but have staff manually review the complaints it ranks as low priority.

WHY IS THIS THE CORRECT ANSWER

A model that systematically disadvantages one group of customers creates unfair outcomes from its first day in use. Finding and fixing the cause before deployment prevents that harm.

AI Best Practices

Intermediate

Question

A colleague wants to paste a client’s confidential contract into a public AI chatbot to get a quick summary. How should you respond?

Select your answer

Suggest they use the company’s approved AI tool, or remove confidential details first, in line with policy.

Suggest they check the company’s AI policy before pasting anything in.

Let them go ahead, but recommend they check the summary carefully for errors.

Let them go ahead, since summaries save time and the chatbot doesn’t keep the data.

WHY IS THIS THE CORRECT ANSWER

Content entered into public AI tools may be stored or used to train models, which can breach client confidentiality and data protection obligations. Pointing the colleague to an approved tool, or to removing sensitive details first, still saves them time while keeping the client’s data protected.

Ready to see Bryq in action?

Start hiring based on real data.

Ready to see Bryq in action?

Start hiring based on

real data.

Ready to see Bryq in action?

Start hiring based on real data.