Ethics, Sustainability and the Future of Risk Management

Introduction

Risk management is not only a technical process for estimating losses. It is also a professional activity involving judgement, power, accountability and social consequences. A bank may comply with a rule yet make an ethically questionable decision; an insurer may price risk accurately yet exclude vulnerable communities; and a financial institution may announce ambitious environmental commitments while continuing practices that contradict them. Contemporary risk management therefore requires both analytical competence and moral responsibility.

Three developments have significantly widened the scope of risk management. First, ethical standards increasingly influence governance, conduct, consumer protection and the legitimacy of financial institutions. Second, climate change and broader environmental, social and governance (ESG) factors create financial risks that may materialise through credit, market, operational, insurance and reputational channels. Third, technology is changing how risks are detected and controlled through RegTech, SupTech, artificial intelligence, real-time analytics and automated decision systems.

This chapter examines these developments and shows how they can be integrated into professional risk practice. The objective is not to treat ethics and sustainability as public-relations subjects, but as dimensions of sound decision-making, resilience and long-term value creation.

Learning Objectives

By the end of this chapter, a student should be able to:

1. Explain the ethical foundations of risk management and apply ethical reasoning to professional decisions.
2. Distinguish legal compliance from ethical conduct and identify conflicts of interest.
3. Explain climate, environmental, social and governance risks and their transmission channels.
4. Assess greenwashing and evaluate the credibility of sustainable-finance claims.
5. Discuss financial inclusion as both a social objective and a risk-management issue.
6. Explain the functions, benefits and limitations of RegTech, SupTech and real-time analytics.
7. Evaluate governance requirements for artificial intelligence and automated risk decisions.
8. Recommend responsible and proportionate risk-management practices.

1. Ethics as a Foundation of Risk Management

1.1 What is professional ethics?

Ethics concerns principles about what ought to be done. Professional ethics applies those principles to decisions made by people who hold specialised knowledge and positions of trust. Banking, finance and insurance professionals influence access to credit, savings, payments, investment opportunities and protection against loss. Their decisions can affect households, businesses, employees, investors and communities.

Ethical risk management asks questions such as:

• Is the decision lawful and consistent with professional duties?
• Who benefits and who bears the potential loss?
• Has material information been disclosed clearly?
• Are customers treated fairly and with due consideration of vulnerability?
• Could the decision create an unacceptable conflict of interest?
• Can the decision be explained and defended to an affected stakeholder?

Ethics is broader than personal honesty. It includes organisational incentives, governance systems, data practices, remuneration structures and the consequences of institutional decisions.

1.2 Compliance and ethics

Compliance means following applicable laws, regulations, standards and internal policies. It is a minimum requirement, not a complete definition of responsible conduct. Ethical behaviour may require an institution to act cautiously even when a practice is technically permissible.

For example, a product disclosure document may satisfy formal requirements but still be excessively complex for ordinary customers. A sales incentive may comply with policy but encourage staff to recommend unsuitable products. A lending model may be statistically accurate overall while producing systematically worse outcomes for a protected or disadvantaged group.

A strong risk culture therefore combines:

• Rules-based controls: policies, limits, approvals and monitoring.
• Principles-based judgement: fairness, honesty, proportionality and stewardship.
• Challenge: willingness to question profitable but harmful decisions.
• Accountability: clear ownership and consequences for misconduct.
• Transparency: records that allow decisions to be reconstructed and reviewed.

1.3 Ethical theories relevant to risk decisions

Several ethical approaches are useful analytical lenses.

Consequentialism evaluates actions by their outcomes. A decision may be justified if it produces the greatest overall benefit, but the approach can risk sacrificing minorities for aggregate gains.

Deontological ethics focuses on duties and principles. Truthfulness, confidentiality, non-discrimination and respect for rights may be obligations even when violating them could generate short-term benefits.

Virtue ethics asks what a responsible professional or institution would do. Qualities such as prudence, integrity, courage and fairness are important where rules do not determine the answer.

Justice and fairness approaches examine how benefits, risks and opportunities are distributed. They are especially relevant to credit access, insurance affordability, algorithmic discrimination and financial inclusion.

In practice, risk professionals often combine these approaches. A decision should be legally permissible, consistent with duties, defensible in its effects and compatible with a culture of integrity.

2. Ethical Decision-Making in Financial Institutions

A practical ethical decision process can be organised into six steps:

1. Define the decision and affected stakeholders. Identify customers, shareholders, employees, regulators, suppliers and communities.
2. Identify facts and uncertainties. Separate verified information from assumptions and disclose material data gaps.
3. Identify duties and constraints. Consider law, regulation, fiduciary duties, contracts, professional standards and internal policy.
4. Identify competing values. Common tensions include profitability versus affordability, privacy versus fraud detection, and speed versus due diligence.
5. Evaluate alternatives and consequences. Consider immediate and long-term effects, including effects on vulnerable groups.
6. Decide, document and review. Record the rationale, controls, escalation route and conditions for reassessment.

A useful test is the explainability and publicity test: could the decision be explained clearly to an affected customer, an independent reviewer, a regulator and the public? This does not replace legal analysis, but it exposes decisions that depend on secrecy or misleading presentation.

Conflicts of interest

A conflict exists when personal, financial or institutional interests could influence professional judgement. Examples include:

• an adviser receiving higher remuneration for selling one product;
• a bank approving a loan to a related party;
• an auditor or consultant reviewing work from which it benefits;
• a risk manager being rewarded only for reducing capital usage;
• a director having an undisclosed interest in a supplier.

Controls include disclosure, recusal, independent review, segregation of duties, remuneration redesign and documented approval. Disclosure alone may be insufficient where the conflict is severe; avoiding or eliminating the conflict may be necessary.

Whistleblowing and speaking up

Ethical risk management depends on employees being able to report misconduct without retaliation. Effective arrangements provide confidential reporting channels, independent investigation, protection against victimisation, board oversight and feedback to reporters. A policy is ineffective if employees reasonably believe that speaking up will damage their careers.

3. Sustainability and ESG Risk

3.1 Meaning of sustainability risk

Sustainability risk is the possibility that environmental, social or governance conditions create financial, operational, legal or reputational effects. ESG factors are not automatically financial risks. They become risk-relevant when they affect cash flows, asset values, liabilities, access to funding, business continuity or stakeholder trust.

The relationship is often described as double materiality:

• Financial materiality: how sustainability issues affect the institution or investment.
• Impact materiality: how the institution or investment affects people and the environment.

A bank may therefore ask both whether climate change could increase loan defaults and whether financed activities contribute materially to environmental harm.

3.2 Environmental risks

Environmental risks include climate change, biodiversity loss, pollution, water scarcity, resource depletion and ecosystem degradation. They may affect financial institutions through several channels:

• Credit risk: lower borrower revenues, higher operating costs or damaged collateral.
• Market risk: repricing of carbon-intensive assets or changes in investor preferences.
• Operational risk: disruption to branches, data centres, transport and suppliers.
• Liquidity risk: rapid withdrawal of funding after environmental controversy.
• Insurance risk: higher frequency or severity of insured events.
• Legal and reputational risk: liability, litigation or loss of stakeholder confidence.

3.3 Climate risk

Climate risk is commonly divided into physical risk and transition risk.

Physical risk arises from climate-related hazards. Acute hazards include floods, storms, wildfires and heatwaves. Chronic hazards include sea-level rise, changing rainfall, prolonged heat and gradual ecosystem changes.

Transition risk arises from the process of moving towards a lower-emission economy. It may result from carbon pricing, regulation, technological change, shifts in consumer demand, litigation or declining demand for carbon-intensive products.

Climate risk has distinctive characteristics:

• its effects may develop over long horizons;
• historical data may not represent future conditions;
• risks can be nonlinear and interconnected;
• outcomes depend on policy and technology pathways;
• the timing of financial loss may differ from the timing of environmental damage.

Climate scenario analysis can explore alternative futures, but it is not a precise forecast. Results depend on assumptions about emissions, economic growth, policy, technology, geography and adaptation.

3.4 Social risks

Social risks include labour exploitation, unsafe working conditions, human-rights violations, discrimination, customer harm, privacy breaches and financial exclusion. They can affect a borrower’s licence to operate, legal exposure, revenue and reputation.

Financial institutions should examine social risks in their own operations and throughout relevant value chains. For example, a project-finance transaction may present human-rights risks even when the direct borrower has strong internal policies, if contractors or suppliers engage in harmful practices.

3.5 Governance risks

Governance risks concern the way organisations are directed and controlled. Relevant factors include board independence, executive accountability, corruption, bribery, tax conduct, data governance, internal controls, remuneration and shareholder rights. Governance weaknesses often amplify other risks: poor oversight can allow environmental misconduct, fraud or unsafe products to persist.

4. Integrating ESG into Risk Management

ESG should not be treated as a separate checklist disconnected from core risk processes. Integration involves five linked activities:

1. Materiality assessment: identify ESG issues relevant to sectors, locations, products and counterparties.
2. Data and due diligence: obtain reliable information, document limitations and challenge unsupported claims.
3. Risk measurement: incorporate relevant factors into credit analysis, valuation, insurance pricing, scenario analysis and operational assessments.
4. Decision and control: use covenants, exclusions, engagement, limits, pricing adjustments, escalation and contingency planning.
5. Monitoring and reporting: track indicators, breaches, outcomes and changes in assumptions.

A simple portfolio emissions-intensity measure may be expressed as:

where I is emissions intensity, E is attributed greenhouse-gas emissions, and V is a defined financial or economic denominator, such as revenue or portfolio value. Units must be stated, for example tonnes of CO₂-equivalent per million currency units of revenue. The measure is meaningful only when the boundary, attribution method, scope of emissions, denominator and reporting period are consistent. It should not be used to compare portfolios blindly when their sectors or data quality differ.

Risk professionals should distinguish measurement from management. A lower intensity does not automatically mean lower absolute emissions or lower transition risk. Metrics require context, targets, governance and credible action plans.

5. Greenwashing and Sustainable-Finance Integrity

Greenwashing occurs when an organisation conveys a misleading impression about the environmental or social characteristics of a product, activity or institution. It may involve false claims, selective disclosure, vague terminology, overstated impact or failure to explain exclusions and trade-offs.

Examples include:

• describing a fund as “green” while holding substantial activities inconsistent with its stated strategy;
• presenting an emissions reduction target without a baseline, timeframe or credible transition plan;
• highlighting a small sustainable project while concealing material environmental harm elsewhere;
• using offsets as if they were equivalent to reducing operational emissions;
• making claims based on estimated data without disclosing uncertainty.

A credible sustainability claim should be:

• specific: it defines the relevant environmental or social attribute;
• measurable: it uses an appropriate metric and baseline;
• comparable: it explains methodology and boundaries;
• substantiated: evidence supports the claim;
• balanced: limitations and adverse impacts are disclosed;
• governed: responsibility for approval, monitoring and correction is clear.

The risk function should challenge marketing language, require evidence, test consistency between product claims and holdings, and escalate material misrepresentation. External assurance can improve confidence but does not transfer responsibility from management.

6. Financial Inclusion and Fairness

Financial inclusion means that individuals and enterprises can access and use useful, affordable and appropriate financial services. Inclusion includes availability, affordability, accessibility, suitability, safety and effective use.

It is relevant to risk management because exclusion can create social and reputational risks, while poorly designed inclusion can create conduct and credit risks. Digital lending, for example, may expand access but also produce excessive borrowing, opaque pricing, privacy violations or discriminatory automated decisions.

Responsible inclusion requires:

• proportionate identification and customer due diligence;
• accessible communication and reasonable assistance;
• assessment of affordability and suitability;
• alternative channels for customers with limited digital access;
• protection of personal data;
• complaint and redress mechanisms;
• monitoring for disparate outcomes.

A model can be accurate on average and still unfair. Institutions should compare approval, pricing, default and complaint outcomes across relevant groups, while recognising that fairness measures can conflict and that sensitive attributes may be needed for legitimate monitoring under appropriate safeguards.

7. RegTech, SupTech and Real-Time Risk Management

7.1 RegTech

RegTech is the use of technology by regulated institutions to meet compliance and risk-management requirements. Applications include automated transaction monitoring, regulatory reporting, identity verification, sanctions screening, document analysis, control testing and policy mapping.

Potential benefits include faster processing, improved consistency, audit trails and the ability to analyse large datasets. Risks include poor data quality, excessive false alerts, vendor concentration, cyber threats, opaque algorithms and automation bias.

7.2 SupTech

SupTech is the use of technology by supervisory authorities. Supervisors may use data analytics, machine learning, natural-language processing and dashboards to identify unusual trends, compare institutions and target examinations. SupTech can improve supervisory efficiency, but authorities must address data confidentiality, model risk, explainability and due process.

7.3 Real-time analytics

Traditional risk reporting often relies on periodic snapshots. Real-time or near-real-time analytics uses continuously updated data to detect events such as unusual payments, liquidity stress, cyber incidents or rapid changes in customer behaviour.

The value of speed depends on data reliability and response capability. An alert that arrives immediately but cannot be investigated, explained or acted upon may increase noise rather than reduce risk. Institutions should define alert ownership, service-level expectations, escalation thresholds and fallback procedures.

7.4 Artificial intelligence and model governance

AI systems can support underwriting, fraud detection, customer service and forecasting. They also introduce model, conduct, privacy, cybersecurity and systemic risks. Good governance includes:

• clearly defined purpose and prohibited uses;
• representative and legally obtained data;
• testing for performance, stability and disparate impacts;
• human oversight appropriate to the decision’s consequences;
• explainability for customers and reviewers;
• access controls, monitoring and change management;
• independent validation and documented limitations;
• a process for correction, appeal and redress.

Human involvement must be meaningful. Merely placing a person at the end of an automated process does not provide genuine oversight if the person lacks authority, time or information to challenge the system.

8. A Future-Oriented Risk Governance Model

A future-ready institution can organise risk governance around four capabilities:

Sensing: collect relevant financial, environmental, social, behavioural and operational signals.

Interpreting: convert data into materiality assessments, scenarios and decisions while recognising uncertainty.

Acting: allocate capital, change products, engage counterparties, limit exposures and protect customers.

Learning: review outcomes, incidents, complaints, model performance and emerging evidence, then revise controls.

This model requires board oversight and clear accountability. The board sets purpose, risk appetite and expectations. Senior management allocates resources and embeds standards. The risk function provides independent challenge. Compliance advises on obligations. Internal audit evaluates governance and controls. Business units own the risks generated by their activities.

Technology should strengthen, not replace, governance. Institutions must avoid “automation washing”—the assumption that a technologically advanced process is automatically objective, accurate or ethical.

9. Illustrative Case: Green Transition Lending and Algorithmic Exclusion

Background

Riverland Bank, a fictional regional bank, launches a sustainability-linked lending programme for small manufacturers. Borrowers receive a lower interest margin if they meet verified energy-efficiency targets. The bank also introduces an automated credit-screening model using financial statements, transaction data and industry information.

Problem

After one year, the programme attracts criticism. The bank’s public report claims that the portfolio supports a just transition, but it does not disclose that many rural firms cannot access the required measurement technology. The credit model also rejects a much higher proportion of applications from firms in remote areas. Management argues that the model is accurate because its overall default prediction is strong.

Analysis

Several risks interact:

1. Greenwashing risk: the sustainability claim may overstate impact because access barriers and exclusions are not disclosed.
2. Conduct risk: small firms may not understand the performance conditions or the consequences of missing targets.
3. Model risk: geographic variables may act as proxies for infrastructure access or socioeconomic disadvantage.
4. Reputational risk: affected firms and community groups may challenge the programme.
5. Credit risk: excluding viable firms may reduce portfolio diversification and relationship knowledge.
6. Data risk: energy data may be incomplete or estimated inconsistently.

The bank should test approval and pricing outcomes by geography and firm characteristics, investigate the causal basis of geographic variables, and provide a human review route. It should offer technical assistance or alternative evidence for firms lacking measurement systems. Sustainability-linked pricing should depend on credible, relevant and verifiable targets, not merely on a borrower’s marketing statement.

Approach and solution

The bank establishes an interdisciplinary committee involving credit, risk, compliance, sustainability, data science and customer representatives. It:

• defines the environmental objective and baseline;
• discloses eligibility rules, exclusions and data limitations;
• validates the model independently and tests outcomes across groups;
• removes or constrains variables lacking a defensible risk rationale;
• introduces manual review for borderline or high-impact decisions;
• funds basic measurement support for eligible small businesses;
• monitors target achievement and unintended exclusion quarterly;
• reports both positive outcomes and adverse effects.

Outcome

The revised programme produces slower initial growth but stronger evidence of environmental impact and fairer access. The bank identifies some borrowers whose transition plans require support rather than rejection. It also discovers that a small number of suppliers had overstated energy savings, leading to revised verification procedures.

Learning points

The case illustrates that sustainability, ethics and technology cannot be managed in isolated departments. A credible programme must align claims, incentives, data, customer outcomes and governance. Accuracy alone does not establish fairness, and a sustainable label does not eliminate the need for ordinary credit and conduct controls.

10. Challenges and Limitations

Contemporary risk management faces several limitations.

Data limitations: ESG data may be incomplete, inconsistent, unaudited or based on estimates. Small enterprises and emerging-market borrowers often have fewer reporting resources.

Uncertainty: Climate pathways, technological changes and social responses cannot be predicted precisely. Scenario analysis should inform resilience, not create false precision.

Measurement disagreement: Different boundaries and methodologies can produce different ESG scores or emissions measures. Users must understand what a metric actually measures.

Short-term incentives: Remuneration and quarterly performance pressure may discourage long-term investment in resilience.

Trade-offs: A decision can improve one objective while worsening another. For example, rapid digitisation may reduce operating costs but exclude people with limited connectivity.

Model limitations: Historical data may encode past discrimination or fail to represent structural change. Models require continuous monitoring and human judgement.

Regulatory fragmentation: Sustainability disclosures, data protection, AI rules and financial regulation may develop at different speeds across jurisdictions.

Reputational overreaction: Institutions may withdraw from high-risk sectors rather than supporting credible transition, thereby shifting activity to less transparent providers. Engagement, financing conditions and transition support may sometimes be better than blanket exclusion.

11. Best-Practice Checklist for Risk Professionals

A responsible risk professional should:

1. Treat ethics, sustainability and technology as core risk matters.
2. Identify who may be harmed or excluded by a decision.
3. Distinguish verified evidence from assumptions and estimates.
4. Challenge incentives that reward volume without customer or societal outcomes.
5. Use clear definitions, boundaries, baselines and time horizons for ESG claims.
6. Integrate material ESG factors into existing risk categories.
7. Test automated decisions for accuracy, stability, explainability and fairness.
8. Maintain human review and accessible appeal mechanisms for consequential decisions.
9. Protect confidential and personal data throughout the information lifecycle.
10. Document decisions, limitations, exceptions and remediation actions.
11. Use scenario analysis to explore uncertainty rather than predict one future.
12. Ensure boards and senior managers receive balanced information, including bad news.
13. Review outcomes, complaints and unintended consequences, not only process compliance.
14. Build staff capability through training in ethics, sustainability, data and technology.

Chapter Summary

Ethical risk management extends beyond legal compliance. It requires fairness, honesty, accountability, transparency and the courage to challenge harmful or misleading decisions. Ethical theories such as consequentialism, duty-based ethics, virtue ethics and justice provide complementary perspectives for analysing difficult choices.

Sustainability risks arise when environmental, social or governance conditions affect financial performance, resilience or stakeholder welfare. Climate risk includes physical and transition risk and can transmit through credit, market, operational, liquidity, insurance, legal and reputational channels. ESG integration requires materiality assessment, reliable data, appropriate metrics, controls, monitoring and disclosure.

Greenwashing undermines sustainable-finance integrity and can create conduct, legal and reputational risk. Financial inclusion must be pursued with affordability, suitability, privacy, accessibility and redress. RegTech, SupTech, real-time analytics and artificial intelligence can improve detection and supervision, but they also create model, data, cyber, privacy and fairness risks.

The future of risk management depends on combining technology with professional judgement and accountable governance. Institutions that sense emerging risks, interpret uncertainty responsibly, act proportionately and learn from outcomes are better positioned to create sustainable and trustworthy financial systems.

Key Terms

Ethics: Principles used to evaluate what ought to be done.

Professional ethics: Duties and standards guiding conduct within a profession.

Risk culture: Shared values, behaviours and incentives that shape risk decisions.

Conflict of interest: A situation in which competing interests may impair impartial judgement.

Sustainability risk: Financial, operational, legal or reputational risk arising from sustainability-related conditions.

ESG: Environmental, social and governance factors relevant to organisational performance and impact.

Double materiality: Assessment of both how sustainability affects an organisation and how the organisation affects society and the environment.

Physical risk: Risk arising from acute or chronic environmental and climate hazards.

Transition risk: Risk arising from adjustment towards a lower-emission or more sustainable economy.

Scenario analysis: Examination of possible future conditions under specified assumptions.

Greenwashing: Misleading communication about environmental or social performance.

Financial inclusion: Access to useful, affordable, appropriate and safe financial services.

RegTech: Technology used by regulated institutions to meet compliance and risk requirements.

SupTech: Technology used by supervisory authorities to improve supervision.

Real-time analytics: Analysis using continuously or rapidly updated data.

Model risk: Risk of adverse consequences from incorrect, misused or poorly governed models.

Algorithmic bias: Systematic and unfair differences in automated outputs or outcomes.

Explainability: The ability to communicate how a decision or model output was produced.

Double materiality: Assessment of outward impacts and inward financial effects.

Transition plan: A structured strategy for moving an organisation or borrower towards defined sustainability objectives.

Review Questions

Short Answer Questions

1. Why is ethical conduct broader than legal compliance?
2. Distinguish physical climate risk from transition risk.
3. What is greenwashing? Give one example.
4. Define RegTech and identify two of its applications.
5. Why is human oversight important in automated risk decisions?

Long Answer Questions

1. Explain how ethical theories can assist a risk professional facing a conflict between profitability and customer fairness.
2. Discuss the main channels through which climate risk can affect a bank or insurer.
3. Explain how ESG factors can be integrated into existing risk-management processes.
4. Evaluate the benefits and limitations of real-time analytics in financial institutions.
5. Describe the governance requirements for responsible use of artificial intelligence in credit or insurance decisions.

Analytical / Application Questions

1. A bank claims that its sustainable loan portfolio is environmentally beneficial, but it discloses neither its baseline nor its exclusions. Analyse the risks and recommend controls.
2. An automated lending model rejects applicants from a particular region at twice the average rate. Its overall accuracy is high. Explain what investigations should be performed before continued use.
3. An insurer faces rapidly increasing flood losses but is under pressure to maintain affordability. Propose an ethically and financially responsible risk-management response.

Multiple Choice Questions with Answers

Question 1

Which statement best distinguishes compliance from ethics?

A. Compliance concerns only environmental matters, while ethics concerns only financial matters.

B. Compliance is following applicable requirements, while ethics also considers fairness, duties and consequences.

C. Compliance is voluntary, while ethics is legally enforceable in every situation.

D. Compliance applies to managers only, while ethics applies to customers only.

Correct Answer: B. Compliance is following applicable requirements, while ethics also considers fairness, duties and consequences.

Question 2

Which of the following is an example of transition risk?

A. A flood damages a bank branch.

B. A heatwave reduces crop yields.

C. A new carbon regulation reduces the value of a carbon-intensive borrower.

D. A storm interrupts an insurer’s call centre.

Correct Answer: C. A new carbon regulation reduces the value of a carbon-intensive borrower.

Question 3

What does double materiality examine?

A. Only the effect of sustainability issues on an organisation’s profits.

B. Only an organisation’s environmental impact.

C. Both the effect of sustainability issues on an organisation and the organisation’s effects on society and the environment.

D. The difference between accounting and market value.

Correct Answer: C. Both the effect of sustainability issues on an organisation and the organisation’s effects on society and the environment.

Question 4

Which practice is most likely to constitute greenwashing?

A. Disclosing a product’s methodology and limitations.

B. Publishing a verified emissions baseline.

C. Describing a fund as sustainable while concealing material holdings inconsistent with that claim.

D. Reporting both positive and negative sustainability outcomes.

Correct Answer: C. Describing a fund as sustainable while concealing material holdings inconsistent with that claim.

Question 5

Which is a primary purpose of financial inclusion?

A. To guarantee every borrower approval regardless of affordability.

B. To provide useful, affordable, appropriate and safe financial services.

C. To eliminate all customer-identification requirements.

D. To replace all physical financial-service channels with digital channels.

Correct Answer: B. To provide useful, affordable, appropriate and safe financial services.

Question 6

RegTech refers primarily to technology used by:

A. Regulated institutions to meet compliance and risk requirements.

B. Only courts to resolve financial disputes.

C. Consumers to compare retail prices.

D. Central banks to set interest rates manually.

Correct Answer: A. Regulated institutions to meet compliance and risk requirements.

Question 7

Why can a highly accurate automated model still create fairness concerns?

A. Accuracy guarantees discrimination.

B. Average accuracy may conceal systematically worse outcomes for a particular group.

C. Models cannot process numerical data.

D. Fairness is unrelated to customer outcomes.

Correct Answer: B. Average accuracy may conceal systematically worse outcomes for a particular group.

Question 8

Which control best supports credible sustainability claims?

A. Using vague terms without defining them.

B. Publishing only favourable results.

C. Defining the baseline, methodology, boundaries, evidence and limitations.

D. Allowing marketing staff to approve claims without review.

Correct Answer: C. Defining the baseline, methodology, boundaries, evidence and limitations.

Question 9

What is a major limitation of climate scenario analysis?

A. It can only use historical weather data and cannot use assumptions.

B. It provides a guaranteed forecast of one future outcome.

C. Its results depend on assumptions about policy, technology, economics and climate pathways.

D. It is unrelated to financial risk.

Correct Answer: C. Its results depend on assumptions about policy, technology, economics and climate pathways.

Question 10

Which governance feature is most important for consequential AI decisions?

A. Removing all human review to improve speed.

B. Allowing the model vendor to make all accountability decisions.

C. Providing meaningful oversight, monitoring, documentation and routes for appeal or correction.

D. Keeping the model’s purpose and limitations undisclosed.

Correct Answer: C. Providing meaningful oversight, monitoring, documentation and routes for appeal or correction.