A regulator asks a bank to describe, in plain terms, how its credit-scoring model reaches individual decisions. Which Responsible AI principle does the request reflect?
Show answer and explanation
Answer: B
Transparency means the behaviour of an AI system can be understood and interrogated by the people affected by it and by those governing it. The obligation is to give meaningful explanations, such as which factors influenced a decision and how strongly, not to disclose proprietary code or confidential records. Explainability tooling, model documentation and clear communication of limitations all support this principle, and it becomes more important as decisions carry greater consequences.
Why the other options are wrong
- A. Open-sourcing algorithms is a licensing and disclosure choice; transparency is satisfied by intelligible explanation, not by public access to implementation.
- C. Training data often contains personal or commercially sensitive records, so publishing it would conflict with privacy rather than demonstrate transparency.
- D. Source code disclosure is not required; what matters is that the reasoning behind outcomes can be explained to stakeholders.
Exam tip. Transparency equals explainability, not open source.