AI-901 - Identify AI concepts and capabilities (42% of the exam) - Section 1.1
Describe the principles of responsible AI and the consideration each addresses in an AI solution.
The six Microsoft responsible-AI principles and what each guards against: fairness, reliability and safety, privacy and security, inclusiveness, transparency, and accountability. The most-tested skill is matching a stated concern to the correct principle. Transparency (making a system and its decisions understandable to the people affected) and accountability (people staying answerable for how the system behaves, with governance and oversight) are the pair most often confused.
fairnessreliability and safetyprivacy and securityinclusivenesstransparencyaccountabilityMicrosoft responsible AI principles
Practice question for this objective
lock_openFree sampleIdentify AI concepts and capabilitieseasy
A retailer is standing up governance for its AI systems and wants to be clear that specific people in the organisation remain answerable for how each system is designed, operated and its outcomes. Which Microsoft responsible AI principle does this describe?
ATransparency
BAccountabilitycheck_circle Correct
CReliability and safety
DFairness
Accountability is the responsible AI principle that people remain answerable for an AI system through governance and oversight. Accountability places responsibility on people: it requires that humans stay answerable for how a system is designed, operated and governed, which distinguishes it from transparency, the principle about making the system understandable.
Why A is wrong: Transparency is the closest distractor because both involve responsible oversight, but transparency is about making the system understandable, whereas accountability is about who answers for it.
Why B is correct: Accountability is the principle that people, not the AI itself, remain answerable for a system through governance and oversight of its design, operation and outcomes, which is exactly the retailer's aim.
Why C is wrong: Reliability and safety is tempting because governance often covers risk, but it concerns dependable and harm-minimising operation, not establishing who is answerable for the system.
Why D is wrong: Fairness is a plausible governance concern, but it addresses equitable treatment of people in a system's outcomes rather than assigning human responsibility and oversight for the system.
Exam traps in Identify AI concepts and capabilities
Answers that look right on this material and are not. Each one is a distractor from a different question in the AI-901 bank for this domain.
Transparency, because stakeholders can understand how the system reaches its outputs
errorWhy it is wrong: Transparency is the principle most often confused with this one, but it is about making the system understandable, whereas assigning owners and governance is about who answers for it.
Transparency
errorWhy it is wrong: A governance framework can involve disclosure, which makes transparency tempting, but transparency is about making the system understandable, not about naming who is answerable for its outcomes.
Transparency, ensuring end users understand how each system produces its outputs
errorWhy it is wrong: Transparency is tempting because governance and understanding often appear together, but it concerns making system behaviour understandable to users, not assigning people who answer for outcomes, so it is not what the framework implements.