WorksheetsPage 1
Total questions: 114
Worksheet time: 57mins
Scenario: HSBC is planning to launch an AI-powered loan approval system. Question: Before implementation, which type of harm should governance professionals assess to avoid negative impacts on individuals and society?
Only economic harms
Reputational, cultural, economic, legal, and regulatory harms
Only technical harms
Only environmental harms
Scenario: An AI system is trained on biased historical hiring data. Question: What individual harm could result?
Increased energy consumption
Unfair treatment and discrimination
Water usage
Lithium extraction
Scenario: HSBC’s AI chatbot is deployed without informing customers they are interacting with AI. Question: Which responsible AI principle is being violated?
Transparency
Fairness
Privacy
Environmental well-being
Scenario: An AI model is so complex that even its developers cannot explain its decisions. Question: What principle is lacking?
Accountability
Explainability
Transparency
Bias mitigation
Scenario: HSBC uses customer social media data for AI training without explicit consent. What privacy harm is most likely?
Overfitting
Appropriation of personal data
Edge case error
Water usage
Scenario: An AI system wrongly attributes a fraudulent transaction to the wrong customer. What is this an example of?
Inference error
Overfitting
Lithium extraction
Energy consumption
Scenario: HSBC’s AI system is trained only on typical cases and fails on rare, unusual transactions. What is this scenario called?
Edge cases and outliers
Overfitting
Underfitting
Transparency
Scenario: An AI model is too simple and misses important patterns in loan applications. Question: What is this called?
Overfitting
Underfitting
Explainability
Lithium extraction
Scenario: HSBC’s AI system is trained on data that is not representative of all customer groups. Question: What harm could result?
Group harm
Environmental harm
Water usage
Energy consumption
Scenario: An AI system is used to influence voting behavior in a national election. Question: What type of harm is this?
Societal harm
Individual harm
Environmental harm
Organizational harm
Scenario: HSBC’s AI model training consumes massive amounts of electricity. Question: What harm is this?
Organizational harm
Environmental harm
Individual harm
Group harm
Scenario: Training a large AI model at HSBC emits over 600,000 pounds of carbon dioxide. Question: What environmental concern does this illustrate?
Water usage
High carbon emissions
Lithium extraction
Data privacy
Scenario: HSBC’s AI system uses lithium batteries, increasing demand for lithium mining. Question: What is a potential harm?
Increased fairness
Environmental strain
Improved transparency
Reduced energy consumption
Scenario: HSBC’s AI system uses water-intensive cooling for its servers. Question: What environmental harm could result?
Water usage
Overfitting
Underfitting
Edge case error
Scenario: HSBC’s AI system is hacked, exposing customer data. Question: What type of harm is this?
Privacy harm
Environmental harm
Societal harm
Scenario: HSBC’s AI system is used to make decisions about employee promotions. Question: What principle should be prioritized to avoid discrimination?
Fairness
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is not regularly audited for accuracy. Question: What harm could result?
Inaccurate decisions
Improved transparency
Reduced energy consumption
Increased fairness
Scenario: HSBC’s AI system is designed to maximize profit without considering customer well-being. Question: Which responsible AI principle is being neglected?
Individual, social and environmental well-being
Transparency
Privacy
Accountability
Scenario: HSBC’s AI system is not understandable to legal teams. Question: What principle is lacking?
Transparency
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system allows customers to challenge decisions made by the system. Question: What principle does this support?
Explainability
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC uses PETs to protect customer data during AI model training. Question: What principle is being supported?
Privacy-enhanced
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be accountable for its decisions. Question: What does accountability mean in this context?
The organization is responsible for AI outcomes
Only developers are responsible
Only users are responsible
No one is responsible
Scenario: HSBC’s AI system is designed to amplify human agency. Question: What principle is this?
Human-centric
Water usage
Scenario: HSBC’s AI system is designed with privacy controls from the start. Question: What approach is this?
Ethics by design
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be fair, transparent, and accountable. Question: Which guidelines are these principles rooted in?
OECD Guidelines and FIPs
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to protect individual rights relative to personal data. Question: Which foundational principle is being followed?
Fair Information Practices (FIPs)
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to minimize negative impacts and maximize positive outcomes. Which framework supports this approach?
NIST AI Risk Management Framework
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be explainable to both technical and non-technical audiences. What principle is being followed?
Transparency
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to allow individuals to opt out of data collection. What principle is being supported?
Privacy and data governance
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be fair to all customer groups. What principle is being followed?
Fairness
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to promote individual, social, and environmental well-being. Question: What principle is being followed?
Individual, social and environmental well-being
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be transparent about its data sources. Question: What principle is being followed?
Transparency
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be accountable for its outputs. Question: What principle is being followed?
Accountability and oversight
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to provide meaningful information about its operation. Question: What principle is being followed?
Transparency and explainability
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to protect data confidentiality using PETs. Question: What principle is being followed?
Privacy-enhanced
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to allow stakeholders to challenge its outputs. Question: What principle is being followed?
Explainability
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to minimize the risk of data misuse. Question: What principle is being followed?
Privacy-enhanced
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be human-centric, accountable, and transparent. Question: What does this combination of principles support?
Trustworthy AI
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be explainable to regulators. Question: What principle is being followed?
Transparency and explainability
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to minimize environmental impact. Question: What principle is being followed?
Individual, social and environmental well-being
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to comply with OECD AI Principles. Question: What does IHTRA stand for in OECD AI Principles?
Inclusive, Human-centric, Transparent, Responsible, Accountable
Individual, Honest, Transparent, Reliable, Accountable
Inclusive, Honest, Transparent, Responsible, Accountable
Inclusive, Human-centric, Trustworthy, Reliable, Accountable
Scenario: HSBC’s AI system is designed to comply with Fair Information Practices. Question: What do FIPs primarily focus on?
Data collection, use, protection, and individual rights
Water usage
Lithium extraction
Scenario: HSBC’s AI system is designed to comply with NIST AI Risk Management Framework. Question: What is risk defined as in this framework?
Probability and magnitude of consequences
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to minimize negative impacts and maximize positive outcomes. Question: What framework supports this approach?
NIST AI Risk Management Framework
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be fair, transparent, and accountable. Question: Which guidelines are these principles rooted in?
OECD Guidelines and FIPs
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to protect individual rights relative to personal data. Question: Which foundational principle is being followed?
Fair Information Practices (FIPs)
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to minimize negative impacts and maximize positive outcomes. Which framework supports this approach?
NIST AI Risk Management Framework
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be explainable to both technical and non-technical audiences. What principle is being followed?
Transparency
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to allow individuals to opt out of data collection. What principle is being supported?
Privacy and data governance
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC’s AI system is designed to be fair to all customer groups. What principle is being followed?
Fairness
Water usage
Lithium extraction
Carbon emissions
Scenario: HSBC is planning to implement AI in its loan approval process. Question: What should be the first step in building AI governance for this initiative?
Select a technology vendor
Understand how the organization operates and its AI maturity
Hire only AI engineers
Ignore stakeholder engagement
Scenario: HSBC is a financial institution considering AI for fraud detection. Question: Which factor should influence its AI governance strategy?
Only company size
Industry/sector and regulatory requirements
Ignore legal requirements
Only product features
Scenario: HSBC wants to launch a generative AI chatbot. Question: Who should be engaged early to strengthen the governance program?
Only developers
Stakeholders across the organization
Only customers
Ignore leadership
HSBC is developing an AI system for internal use. What should the strategy include to inform leadership?
Define roles and responsibilities
Ignore organizational preferences
Only focus on technical details
Avoid documentation
HSBC is evaluating a third-party AI tool for customer service. What is a critical step in AI procurement?
Ignore vetting process
Assess jurisdictional compliance and purpose
Only check price
Skip risk assessment
HSBC is deploying a high-risk AI system. What must deployers ensure?
Input data is relevant, representative, error-free, and complete
Ignore data quality
Only monitor outputs
Avoid staff training
HSBC is using an AI system to automate recruitment. Who is responsible for ensuring safety, transparency, and accountability before market release?
Providers
Users
Scenario: HSBC’s AI system makes consequential decisions about customers. Question: What must deployers do?
Notify consumers when high-risk AI is used
Hide AI usage
Only notify regulators
Ignore impact assessments
Scenario: HSBC is using a generative AI tool for creative work. Question: What is a key responsibility for users?
Recognize when engaging with AI
Ignore guidelines
Only provide feedback
Avoid exercising rights
Scenario: HSBC is tailoring its AI governance model. Question: What factor should be considered?
Company size and resource availability
Ignore maturity
Only focus on technology
Avoid industry standards
Scenario: HSBC is a small company integrating AI. Question: How might it approach AI governance?
Adopt a lightweight framework scaled to its size and resources
Implement a full enterprise-grade governance identical to large corporations
Avoid formal governance and rely on ad hoc decisions
Outsource all accountability to AI vendors without internal oversight
Scenario: HSBC is a large multinational deploying AI. Question: What governance structure might it use?
Create new AI-specific offices and oversight
Ignore detailed processes
Only use existing privacy teams
Avoid documentation
Scenario: HSBC is assessing risk tolerance for a new AI project. Question: What should influence the decision to use AI?
Risk assessment of the use case
Ignore risk
Only consider cost
Avoid stakeholder input
Scenario: HSBC is developing an AI governance policy. Question: What should the policy ensure?
Oversight and accountability across the AI life cycle
Only focus on deployment
Ignore training
Avoid impact assessments
Scenario: HSBC is conducting a use case assessment for a new AI initiative. When should this assessment be performed?
Before implementation and throughout the AI life cycle
Only after deployment
Ignore regulatory compliance
Only for low-risk systems
Scenario: HSBC is building an AI governance framework. What is a key principle?
Align AI with stakeholder objectives and compliance
Ignore bias
Only focus on innovation
Avoid ethical considerations
Scenario: HSBC is choosing an AI governance model. What is a characteristic of a centralized model?
One team/person is responsible for all AI affairs
Decision-making is delegated to lower levels
Combination of central and local entities
No oversight
Scenario: HSBC is considering a decentralized governance model. What is a key feature?
Delegated authority and bottom-up decision-making
Only top-down control
No stakeholder engagement
Ignore local entities
Scenario: HSBC is implementing a hybrid governance model. What does this involve?
Central entity with local support for policies
Only central control
No local involvement
Ignore policy directives
Scenario: HSBC is defining roles for AI governance. Who should be included?
Researchers, data scientists, AI/ML engineers, non-AI engineers
Only developers
Only legal teams
Ignore operations
Scenario: HSBC is forming an AI review committee. What is a benefit of internal committees?
Focused governance activities
Only external members
Ignore ethics
Avoid stakeholder input
Scenario: HSBC is engaging external stakeholders in governance. What is a benefit?
Broader membership and perspectives
Only internal views
Ignore external input
Avoid transparency
Scenario: HSBC is training staff on AI systems. Question: What should training include?
Purpose, limitations, security, and privacy controls
Only technical details
Ignore human impacts
Avoid AI literacy
Scenario: HSBC is rolling out generative AI tools to employees. Question: What guidance should be provided?
Avoid sharing sensitive or classified information without approval
Ignore privacy risks
Only focus on creativity
Avoid training
Scenario: HSBC is improving AI literacy among staff. Question: Why is AI literacy important?
Enables responsible, ethical, and effective engagement with AI
Only for technical teams
Ignore governance
Avoid training
Scenario: HSBC is operationalizing responsible AI. Question: What is a key step?
Establish clear technical standards and runbooks
Only focus on deployment
Ignore organizational rules
Avoid legal updates
Scenario: HSBC is updating its legal structures for AI. Question: Why is this necessary?
Reflect new roles and responsibilities
Only for compliance
Ignore legal risks
Avoid documentation
Scenario: HSBC is fostering a culture of responsible AI. Question: What is a key incentive?
Highlight customer value and trust
Only focus on profit
Ignore diversity
Avoid training
Scenario: HSBC is promoting diversity in AI governance. Question: Why is this important?
Ensures inclusivity and avoids disadvantaging groups
Only for compliance
Ignore cultural variations
Avoid policy review
Scenario: HSBC is defining responsible AI as a discipline. Question: What is a benefit?
Reinforces AI’s value and supports governance community
Only for HR
Ignore success measures
Avoid rewarding practitioners
Scenario: HSBC is setting common AI terms and taxonomy. Question: What is a benefit?
Improves clarity and shared understanding across the organization
Only for compliance documentation
Ignore differing definitions across teams
Avoid standard references
Scenario: HSBC is providing knowledge resources and training. Question: What is the goal?
Promote ethical behaviour in AI practices
Only for compliance
Ignore ongoing education
Avoid resource allocation
Scenario: HSBC is embedding trustworthy AI in its operating model. Question: How is this achieved?
Practice responsible AI processes and risk management
Only focus on technology
Ignore privacy
Avoid accountability
Scenario: HSBC is scaling its AI operations. Question: What must be ensured?
AI can withstand increased users, scale, and data
Only focus on deployment
Ignore operational challenges
Avoid risk management
Scenario: HSBC is validating its AI systems. Question: What is a key requirement?
Confirm AI systems are safe and secure
Only focus on accuracy
Scenario: HSBC is ensuring AI integrity. Question: What principles should be applied?
Transparency, explainability, fairness, and non-discrimination
Only technical standards
Ignore human oversight
Avoid documentation
Scenario: HSBC is enabling human oversight in AI systems. Question: What is the benefit?
Promotes human values and accountability
Only for compliance
Ignore oversight
Avoid human involvement
Scenario: HSBC is selecting an AI governance framework. Question: What factors should be considered?
Principles, risk tolerance, jurisdiction, industry, business strategy, purpose, size
Only technology
Ignore compliance
Avoid risk assessment
Scenario: HSBC is developing AI for external sales. Question: What must be addressed?
User base, jurisdictional requirements, maintenance, and monitoring
Only focus on sales
Scenario: HSBC is procuring a third-party AI system. Question: What is a key governance step?
Vetting process and jurisdictional compliance
Only check price
Ignore risk assessment
Avoid purpose review
Scenario: HSBC is auditing AI system usage. Question: What should be included?
Access controls, auditing process, alignment with principles, regulatory compliance
Only technical logs
Ignore regulatory requirements
Avoid impact assessments
Scenario: HSBC is updating its risk management program for AI. Question: What is a key consideration?
Align new AI risk strategies with existing frameworks
Only focus on new risks
Ignore existing programs
Avoid integration
Scenario: HSBC is assessing operational risks of AI. Question: What is a common operational risk?
High costs for hardware, data, and skilled professionals
Only focus on software
HSBC is mitigating legal risks in AI deployment. What is a key mitigant?
Comprehensive governance frameworks and legal reviews
Only focus on technology
Ignore compliance
Avoid documentation
HSBC is addressing security risks in AI. What is a key threat?
Adversarial attacks and data poisoning
Only focus on accuracy
Ignore security
Avoid encryption
HSBC is mitigating privacy risks in AI. What practice should be adopted?
Data minimization and transparency
Only focus on data collection
Ignore informed consent
Avoid privacy regulations
HSBC is addressing business risks in AI. What is a key risk?
Bias and discrimination from poor data quality
Only focus on profit
Ignore job displacement
Avoid vendor evaluation
Scenario: HSBC is mitigating job displacement due to AI. What strategy can help?
Reskilling and upskilling employees
Only focus on automation
Ignore social impacts
Avoid training
Scenario: HSBC is evaluating vendor dependence for AI systems. What is a risk?
Vendor lock-in and failure
Only focus on price
Ignore contingency planning
Avoid vendor evaluation
Scenario: HSBC is conducting an AI system impact assessment (AIIA). What key areas should be covered?
Privacy, bias, transparency, accountability, security, broad impacts
Only technical details
Ignore human rights
Avoid documentation
Scenario: HSBC is deploying an AI system for credit scoring in the EU. Question: What regulatory approach will apply to this high-risk system?
Minimal oversight
Strict obligations including risk management and conformity assessment
No documentation required
Only voluntary guidelines
Scenario: HSBC is considering using an AI-powered chatbot for customer service. Question: What is the likely regulatory requirement for this limited-risk application?
Ban on use
Disclosure to users that they are interacting with AI
Mandatory impact assessment
No requirements
Scenario: HSBC is importing an AI system from the US to the EU. Question: What must the importer ensure before market entry?
The system complies with EU laws
Only US laws apply
No compliance needed
Ignore documentation
Scenario: HSBC is distributing an AI system in the supply chain. Question: What is a key responsibility for distributors?
Ensure conformity and proper handling
Ignore compliance
Only focus on marketing
No obligations
Scenario: HSBC is developing a general-purpose AI model (GPAI) for multiple applications. Question: What is a common global obligation for GPAI providers?
Maintain technical documentation and transparency
No documentation needed
Only focus on deployment
Ignore downstream risks
Scenario: HSBC is deploying a facial recognition system in public spaces. Question: What risk level does this system fall under in the EU AI Act?
Minimal risk
Prohibited or unacceptable risk
Limited risk
No risk
Scenario: HSBC is using AI for entertainment purposes, such as music generation. Question: What is the regulatory approach for this minimal-risk application?
Minimal oversight with voluntary codes of conduct
Full conformity assessment and CE marking
Mandatory comprehensive risk management and post-market monitoring
Prohibited under the EU AI Act
Scenario: HSBC is deploying an AI system for medical device diagnostics. Question: What obligations apply to this high-risk system?
Risk management, data governance, technical documentation, human oversight
No obligations
Only voluntary guidelines
Ignore data quality
Scenario: HSBC is using a generative AI model to create deepfake images. Question: What must be disclosed under most global regulations?
That content is AI-generated or manipulated
No disclosure needed
Only disclose to regulators
Ignore user information
Scenario: HSBC is deploying AI for employment decisions. Question: What risk level and obligations apply?
High-risk; strict obligations including impact assessment and transparency
Minimal risk; voluntary standards
No obligations
Only focus on entertainment
Scenario: HSBC is using AI for predictive policing. Question: What is the regulatory stance in the EU?
Prohibited or unacceptable risk
Minimal risk
Limited risk
No risk
Scenario: HSBC is deploying a GPAI model with systemic risk. Question: What additional obligations apply?
Red-teaming, incident reporting, robust cybersecurity, energy consumption disclosure
No additional obligations
Only documentation
Ignore risk management
Scenario: HSBC is launching an AI system in South Korea. Question: What must a foreign provider do?
Designate a domestic agent for compliance reporting
Ignore local laws
Only comply with home country laws
No obligations
Scenario: HSBC is deploying AI in healthcare in South Korea. Question: What must business operators review in advance?
Whether the AI is high impact
Only focus on entertainment
Ignore risk assessment
No review needed
