Practise AI governance language: model registry, approval workflows, AI governance committees, and responsible deployment communication.
0 / 22 completed
1 / 22
A 'model registry' in an MLOps governance framework is used to:
A model registry (MLflow, SageMaker Model Registry) tracks model versions, evaluation results, approval status, and deployment history — providing auditability.
2 / 22
An AI governance committee is responsible for:
AI governance committees (or ethics boards) review proposed AI applications against organisational policies, legal requirements, and ethical standards before approving deployment.
3 / 22
In the context of AI deployment, 'human-in-the-loop' (HITL) means:
HITL keeps humans in the decision loop — the AI provides recommendations or flags cases, but a human validates before the decision is acted upon. Critical for high-stakes use cases.
4 / 22
Which phrase best describes an AI system's 'kill switch' or 'human override'?
The EU AI Act and AI safety frameworks require high-risk AI systems to include mechanisms for human override — allowing authorised personnel to stop or correct the system.
5 / 22
What does 'model explainability' provide to stakeholders?
Explainability enables affected people to understand AI decisions — required by GDPR (right to explanation) for automated decisions and by AI governance frameworks for high-risk use cases.
6 / 22
PR Description
During a code review of a new fraud detection model, Sarah notices the following comment from David:
"This model uses a complex LSTM architecture. We need to ensure it's thoroughly tested for adversarial attacks – specifically, inputs designed to mislead the network. Consider adding more detailed documentation on the attack vectors we've considered and the mitigation strategies employed."
This is a good example of proactive governance communication within a code review. David's comment isn't simply pointing out a technical concern; he's specifically requesting documentation on attack vectors and mitigation strategies – crucial elements for demonstrating responsible AI development and meeting regulatory requirements. The incorrect options misinterpret the comment as insufficient or overly cautious, failing to recognize its value in establishing accountability and risk management around model deployment.
7 / 22
During a standup update, Alex is explaining the deployment of their new AI-powered customer support chatbot. He mentions that 'we have a human-in-the-loop system' in place. Which of the following best describes what this means within the context of AI governance?
Alex: 'We have a human-in-the-loop system.'
This question focuses on HITL implementation. Option A describes monitoring and escalation, not the core of a human-in-the-loop system. Option B details automated flagging but doesn't involve direct human intervention in correcting responses. Option C accurately represents the real-time feedback loop, allowing for continuous improvement and ensuring alignment with governance principles. Option D is an incorrect interpretation of paused operation.
8 / 22
As a lead engineer, you're reviewing the PR description for a model that predicts loan defaults. The description states: 'The model incorporates feature selection to improve performance and reduce bias.' Which of the following is the MOST relevant governance consideration stemming from this statement?
PR Description: 'The model incorporates feature selection to improve performance and reduce bias.'
This question tests understanding of bias mitigation. While performance is important, documentation of feature selection – especially regarding bias – is paramount. Option A prioritizes speed over governance. Option B directly addresses the need for transparency and accountability related to feature selection and bias. Options C and D represent flawed reasoning about model evaluation and risk acceptance, respectively.
9 / 22
PR Description
During a code review of a new fraud detection model, Sarah notices the following comment from David:
"This model uses a complex LSTM architecture. We need to ensure it's thoroughly tested for adversarial attacks – specifically, inputs designed to mislead the network. Consider adding more detailed documentation on the attack vectors we've considered and the mitigation strategies employed."
This is a good example of proactive governance communication within a code review. David's comment isn't simply pointing out a technical concern; he's specifically requesting documentation on attack vectors and mitigation strategies – crucial elements for demonstrating responsible AI development and meeting regulatory requirements. The incorrect options misinterpret the comment as insufficient or overly cautious, failing to recognize its value in establishing accountability and risk management around model deployment.
10 / 22
During a standup update, Alex is explaining the deployment of their new AI-powered customer support chatbot. He mentions that 'we have a human-in-the-loop system' in place. Which of the following best describes what this means within the context of AI governance?
Alex: 'We have a human-in-the-loop system.'
This question focuses on HITL implementation. Option A describes monitoring and escalation, not the core of a human-in-the-loop system. Option B details automated flagging but doesn't involve direct human intervention in correcting responses. Option C accurately represents the real-time feedback loop, allowing for continuous improvement and ensuring alignment with governance principles. Option D is an incorrect interpretation of paused operation.
11 / 22
As a lead engineer, you're reviewing the PR description for a model that predicts loan defaults. The description states: 'The model incorporates feature selection to improve performance and reduce bias.' Which of the following is the MOST relevant governance consideration stemming from this statement?
PR Description: 'The model incorporates feature selection to improve performance and reduce bias.'
This question tests understanding of bias mitigation. While performance is important, documentation of feature selection – especially regarding bias – is paramount. Option A prioritizes speed over governance. Option B directly addresses the need for transparency and accountability related to feature selection and bias. Options C and D represent flawed reasoning about model evaluation and risk acceptance, respectively.
12 / 22
PR Description
During a code review of a new fraud detection model, Sarah notices the following comment from David:
"This model uses a complex LSTM architecture. We need to ensure it's thoroughly tested for adversarial attacks – specifically, inputs designed to mislead the network. Consider adding more detailed documentation on the attack vectors we've considered and the mitigation strategies employed."
This is a good example of proactive governance communication within a code review. David's comment isn't simply pointing out a technical concern; he's specifically requesting documentation on attack vectors and mitigation strategies – crucial elements for demonstrating responsible AI development and meeting regulatory requirements. The incorrect options misinterpret the comment as insufficient or overly cautious, failing to recognize its value in establishing accountability and risk management around model deployment.
13 / 22
During a standup update, Alex is explaining the deployment of their new AI-powered customer support chatbot. He mentions that 'we have a human-in-the-loop system' in place. Which of the following best describes what this means within the context of AI governance?
Alex: 'We have a human-in-the-loop system.'
This question focuses on HITL implementation. Option A describes monitoring and escalation, not the core of a human-in-the-loop system. Option B details automated flagging but doesn't involve direct human intervention in correcting responses. Option C accurately represents the real-time feedback loop, allowing for continuous improvement and ensuring alignment with governance principles. Option D is an incorrect interpretation of paused operation.
14 / 22
As a lead engineer, you're reviewing the PR description for a model that predicts loan defaults. The description states: 'The model incorporates feature selection to improve performance and reduce bias.' Which of the following is the MOST relevant governance consideration stemming from this statement?
PR Description: 'The model incorporates feature selection to improve performance and reduce bias.'
This question tests understanding of bias mitigation. While performance is important, documentation of feature selection – especially regarding bias – is paramount. Option A prioritizes speed over governance. Option B directly addresses the need for transparency and accountability related to feature selection and bias. Options C and D represent flawed reasoning about model evaluation and risk acceptance, respectively.
15 / 22
PR Description
During a code review of a new fraud detection model, Sarah notices the following comment from David:
"This model uses a complex LSTM architecture. We need to ensure it's thoroughly tested for adversarial attacks – specifically, inputs designed to mislead the network. Consider adding more detailed documentation on the attack vectors we've considered and the mitigation strategies employed."
This is a good example of proactive governance communication within a code review. David's comment isn't simply pointing out a technical concern; he's specifically requesting documentation on attack vectors and mitigation strategies – crucial elements for demonstrating responsible AI development and meeting regulatory requirements. The incorrect options misinterpret the comment as insufficient or overly cautious, failing to recognize its value in establishing accountability and risk management around model deployment.
16 / 22
During a standup update, Alex is explaining the deployment of their new AI-powered customer support chatbot. He mentions that 'we have a human-in-the-loop system' in place. Which of the following best describes what this means within the context of AI governance?
Alex: 'We have a human-in-the-loop system.'
This question focuses on HITL implementation. Option A describes monitoring and escalation, not the core of a human-in-the-loop system. Option B details automated flagging but doesn't involve direct human intervention in correcting responses. Option C accurately represents the real-time feedback loop, allowing for continuous improvement and ensuring alignment with governance principles. Option D is an incorrect interpretation of paused operation.
17 / 22
As a lead engineer, you're reviewing the PR description for a model that predicts loan defaults. The description states: 'The model incorporates feature selection to improve performance and reduce bias.' Which of the following is the MOST relevant governance consideration stemming from this statement?
PR Description: 'The model incorporates feature selection to improve performance and reduce bias.'
This question tests understanding of bias mitigation. While performance is important, documentation of feature selection – especially regarding bias – is paramount. Option A prioritizes speed over governance. Option B directly addresses the need for transparency and accountability related to feature selection and bias. Options C and D represent flawed reasoning about model evaluation and risk acceptance, respectively.
18 / 22
David left this comment on a PR for an AI-powered sentiment analysis model: "To mitigate potential bias, we've implemented differential privacy techniques during training. This adds noise to the data to protect individual user information.". Which of the following best explains differential privacy in this context?
Differential privacy is a technique designed to protect individual privacy while still allowing for useful statistical analysis. The key idea is adding carefully calibrated noise to calculations – in this case, during training – so that the presence or absence of any *one* user's data has a minimal impact on the model's learned parameters. Option A is incorrect because it implies perfect accuracy, which isn't achievable with privacy-preserving techniques. Option D describes anonymization, not differential privacy.
19 / 22
Maria sends a Slack message to the team after deploying an AI fraud detection system: 'We've implemented a real-time monitoring dashboard and a 'kill switch' – if we see unusual patterns or high false positive rates, we can immediately halt the model's operation.' What does the 'kill switch' primarily represent?
The 'kill switch' is a critical safety feature that provides human oversight and control over an AI system. It's a manual intervention point – allowing operators to quickly stop the model if it malfunctions or exhibits undesirable behavior. Option A describes anomaly detection; option C refers to retraining; and option B mischaracterizes its purpose.
20 / 22
PR Description:
"The model utilizes SHAP values to explain individual predictions. This allows us to understand which features are most influential in driving a specific prediction and identify potential areas for bias mitigation."
What is the primary purpose of using SHAP values in this PR description?
SHAP (Shapley Additive Explanations) values provide a way to understand how each feature contributes to a model's prediction. This is crucial for explaining *why* a model made a certain decision, which directly relates to interpretability and identifying potential biases. Option A misrepresents SHAP's function; option C focuses on hyperparameter optimization; and option B describes visualization.
21 / 22
Ben is giving a stand-up update: 'We're deploying our new AI chatbot for customer support. We've incorporated a human-in-the-loop system where agents can step in and take over the conversation if the bot encounters complex or sensitive issues.' What does 'human-in-the-loop' mean in this scenario?
'Human-in-the-loop' signifies that humans retain oversight and control within an automated system. In this case, it means human agents are *monitoring* the chatbot's interactions, ready to step in if the AI struggles or encounters situations outside its capabilities – not that the bot is entirely independent. Option A describes a consultive model; option C accurately defines intervention; and option B is overly simplistic.
22 / 22
PR Description:
"The model uses Monte Carlo Tree Search (MCTS) to handle uncertainty in its predictions. This allows it to explore multiple possible outcomes and select the most likely one based on a probabilistic assessment."
MCTS is a search algorithm frequently used in AI systems (especially game playing) to handle uncertainty. The key benefit here is that it allows the model to *quantify* its confidence – providing an estimate of how certain it is about each prediction. This insight into potential uncertainty is crucial for decision-making and risk assessment. Option A is incorrect; option B describes a computational optimization technique, and option C misrepresents the algorithm's function.
What will I practice in "AI Governance Communication"?
This is an AI Ethics exercise set. It walks through 22 scenario-based multiple-choice questions built around real usage of AI Ethics terminology that IT professionals encounter on the job.
Is this exercise free to use?
Yes. Every exercise on CoderSlingo, including this one, is free to complete with no account, sign-up, or paywall.
How many questions are in this exercise?
This set contains 22 questions. Each one shows immediate feedback and a detailed explanation after you answer, so you learn the correct usage right away rather than waiting for a final score.
Do I need prior experience to complete this exercise?
No prior experience is required. Each question includes a full explanation covering the reasoning behind the correct answer, so the exercise itself teaches the AI Ethics vocabulary as you go.
Can I retry the exercise if I get questions wrong?
Yes — use the "Try again" button on the results screen to reset your answers and go through all the questions again. There is no limit on attempts.
Is my progress saved?
Your answers and score for the current session are tracked in the browser as you go. No account or login is needed, and there is nothing to install.
What if I don't understand a term used in a question?
Read the explanation shown after you answer each question — it breaks down the correct term in plain English with a real-world example. You can also check the site Glossary for quick definitions.
How is this different from reading a blog article on the topic?
Exercises like this one are interactive drills that test and reinforce specific vocabulary through multiple-choice questions, while blog articles explain concepts in prose. Practising here after reading builds active recall, not just passive recognition.
Where can I find more AI Ethics exercises?
See the AI Ethics exercises hub for the full set of related pages, or browse all exercise categories from the main Exercises index.
Can I use this exercise to prepare for a technical interview?
Yes — AI Ethics vocabulary comes up often in technical discussions and interviews. Pair this exercise with our dedicated Interview Preparation section for role-specific practice.