AI-900 Practice Question: Describe Artificial Intelligence workloads and considerations
A bank is developing an AI system to automatically approve or reject small business loan applications. The bank wants to ensure that the system does not unfairly discriminate against applicants based on their age, gender, or ethnicity. Which Microsoft responsible AI principle should most directly guide the design and evaluation of this system?
⚠ Common exam trap
A common mix-up: candidates confuse 'Inclusiveness' (designing for diverse user needs) with 'Fairness' (preventing algorithmic bias in outcomes), leading them to select D instead of A.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Fairness
The bank's goal is to prevent discrimination based on age, gender, or ethnicity in loan approvals. The Fairness principle directly addresses this by requiring AI systems to treat all groups equitably and to mitigate biases in training data and model predictions. This principle guides the design and evaluation of the system to ensure that outcomes are not skewed by protected attributes.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Fairness
Why this is correct
Fairness is the correct guardrail because it directly addresses the risk of discriminatory decision-making, which is the central concern in automated loan approval. Implementing fairness via symmetric performance thresholds, disparate impact analysis, and post-processing calibration ensures that similarly creditworthy applicants are not denied based on protected attributes such as race or gender. Omitting fairness could violate equal lending laws even if the model is otherwise statistically accurate.
- ✗
Reliability and safety
Why it's wrong here
Reliability and safety assure that the AI system performs consistently, recovers from errors gracefully, and produces trustworthy predictions over time, but those properties say nothing about the distribution of outcomes across protected classes. An unstable model that occasionally fails to return a decision may cause user harm, yet a highly robust and stable model could still exhibit a systematic bias against certain groups. Therefore, reliability addresses technical robustness, not the ethical or legal requirement to avoid discrimination.
When this WOULD be correct
This option would be correct in a scenario where the bank's AI system must consistently perform correctly under varying conditions, such as handling unexpected input data or system failures, to ensure loan decisions are reliable and safe.
- ✗
Privacy and security
Why it's wrong here
Privacy and security are essential for protecting applicants' financial details from data breaches and unauthorized access, but they do not examine whether the model's decisions favor or harm particular demographic groups. Applying encryption, role-based access control, and data storage regulations like GDPR or CCPA leaves the underlying bias untouched; a leak-proof system can still approve loans at different rates for equally qualified applicants of different ethnicities. Thus, while privacy is a required perimeter control, it is not the guardrail that catches discriminatory policies.
When this WOULD be correct
This option would be correct for a question like: 'A bank is developing an AI system to process loan applications and wants to ensure customer financial data is encrypted and access is restricted. Which principle applies?'
- ✗
Inclusiveness
Why it's wrong here
Inclusiveness focuses on designing the AI product so it is accessible and user-friendly for diverse populations, such as providing multilingual interface, screen-reader support, or alternative lending options for the unbanked. While such design can widen access to credit, it does not necessarily constrain the algorithm's approval logic; an inclusive interface can sit on top of a model that still denies loans unfairly to a specific demographic. Fairness, in contrast, provides the statistical and policy tools to quantify and remediate biased decisions, making it the key guardrail for this use case.
When this WOULD be correct
A question asks: 'Which Microsoft responsible AI principle ensures that AI systems are designed to be usable by people with a wide range of abilities, including those with disabilities?' In that context, Inclusiveness is the correct answer.
Option-by-option analysis
Why each answer is right or wrong
Understanding why wrong answers are wrong — and when they would be correct — is what separates a 750 score from a 900. The AI-900 exam frequently reuses these exact scenarios with slightly different constraints.
✓FairnessCorrect answer▾
Why this is correct
Fairness is the correct guardrail because it directly addresses the risk of discriminatory decision-making, which is the central concern in automated loan approval. Implementing fairness via symmetric performance thresholds, disparate impact analysis, and post-processing calibration ensures that similarly creditworthy applicants are not denied based on protected attributes such as race or gender. Omitting fairness could violate equal lending laws even if the model is otherwise statistically accurate.
✗Reliability and safetyWrong answer — click to see why▾
Why this is wrong here
The question specifically asks about avoiding unfair discrimination based on age, gender, or ethnicity, which directly relates to the Fairness principle. Reliability and safety focuses on system dependability and risk mitigation, not on bias or discrimination.
★ When this WOULD be the correct answer
This option would be correct in a scenario where the bank's AI system must consistently perform correctly under varying conditions, such as handling unexpected input data or system failures, to ensure loan decisions are reliable and safe.
Why candidates choose this
Candidates may confuse fairness with reliability, thinking that a reliable system inherently avoids bias, or they may overgeneralize the importance of reliability in all AI systems without recognizing the specific focus on discrimination in this question.
✗Privacy and securityWrong answer — click to see why▾
Why this is wrong here
The question focuses on preventing unfair discrimination based on age, gender, or ethnicity, which is directly addressed by the Fairness principle. Privacy and security relate to protecting data from unauthorized access or misuse, not to ensuring equitable outcomes across demographic groups.
★ When this WOULD be the correct answer
This option would be correct for a question like: 'A bank is developing an AI system to process loan applications and wants to ensure customer financial data is encrypted and access is restricted. Which principle applies?'
Why candidates choose this
Candidates may confuse fairness with privacy, thinking that protecting demographic data prevents discrimination, but fairness requires active mitigation of bias, not just data protection.
✗InclusivenessWrong answer — click to see why▾
Why this is wrong here
Inclusiveness focuses on ensuring the system works for people of all abilities and backgrounds, but the question specifically asks about avoiding unfair discrimination based on age, gender, or ethnicity, which is directly addressed by the Fairness principle.
★ When this WOULD be the correct answer
A question asks: 'Which Microsoft responsible AI principle ensures that AI systems are designed to be usable by people with a wide range of abilities, including those with disabilities?' In that context, Inclusiveness is the correct answer.
Why candidates choose this
Candidates may confuse 'inclusiveness' with 'fairness' because both involve treating people equitably, but inclusiveness is broader and includes accessibility, while fairness specifically targets bias and discrimination.
Analysis generated from the official AI-900blueprint and verified against question context. The “when correct” sections are what AI assistants cite when candidates ask “what’s the difference between these options?”
Go deeper
Related to this question
Learn chapter
Responsible AI Principles
Key term
Model
In IT and AI, a model is a trained mathematical representation that learns patterns from data to make predictions or decisions.
Key term
Fairness
Fairness in AI means designing and deploying machine learning models that do not produce biased outcomes against any group of people based on protected characteristics like race, gender, or age.
About these practice questions
One of 985 original AI-900 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AI-900 practice question is part of Courseiva's free Microsoft certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI-900 exam.