A hiring model scores one group lower than others with equal qualifications. Which principle is at risk?
Why D is the answer
Fairness
Fairness means treating similar people equally. Biased outcomes are its clearest failure.
In detail
- Equal qualifications should get equal scores.
- A gap between groups is a fairness problem.
- Test outcomes across groups before release.
Use case
An illustrative scenario, not a real incident.
A hiring model scores one group lower despite equal qualifications. The team tests results by group and retrains on better data.
See it in different ways
mindmap
root((Responsible AI))
Fairness
Reliability and safety
Privacy and security
Inclusiveness
Transparency
Accountabilitymindmap
root((Responsible AI))
Fairness
Treats groups equally
Reliability and safety
Consistent, fails safely
Privacy and security
Protects data
Inclusiveness
Works for everyone
Transparency
Can be explained
Accountability
People are answerableflowchart TD
Q{"Do equal candidates get equal scores?"}
Q -->|"Yes"| A0["No sign of bias here"]
Q -->|"No"| A1["Fairness problem, so investigate"]Key point: Unequal treatment is a fairness problem.
Why the other options are wrong
- A. Transparency about how the data was stored
- The issue is the outcome, not how data was stored.
- B. Reliability of the hosting service
- The hosting service is not the concern.
- C. Inclusiveness in the design of the interface
- The interface is not the concern.