Learning Objectives:

  • Identify the risks of AI adoption in banking.

  • Explain ethical issues in AI, including bias and transparency.

  • Understand the regulatory landscape for AI in finance.

7.1 AI Risks in Banking

The NUS programme includes a full day on AI risk management, covering :

  • Emerging Cybercrime Risks: Adversarial AI, deepfake fraud, phishing, synthetic identities.

  • Operational Risks: Model failure, data breaches.

  • Systemic Risks: Concentration of AI models leading to market instability.

7.2 Ethical Issues in AI

The NUS course covers AI ethical issues, including :

  • Transparency and Black Box Issues: Many AI models are difficult to interpret.

  • AI Biases: Models may perpetuate or amplify existing biases.

  • Data Privacy: Concerns about the use of personal data.

  • Societal Implications: Job displacement and inequality.

7.3 Regulatory Frameworks for AI

The HKU SPACE course covers regulations, compliance, and ethical considerations in AI :

  • European Union Artificial Intelligence Act: The landmark EU regulation for AI.

  • HKMA Guideline on Generative AI: Hong Kong’s approach.

  • Biases and Fairness: Ensuring AI systems are fair.

  • Model Validation: Ensuring AI models are reliable.

  • Ethical Considerations: Broader ethical issues surrounding AI in banking.

The UAE’s “Principles of AI and its Applications in Banking” programme covers current and upcoming AI regulations, including strategies, frameworks, and tools for managing AI risks .

7.4 AI Governance and Best Practices

Managing AI risk requires robust governance:

  • Model Validation: Independent validation of AI models.

  • Explainability: Using Explainable AI (XAI) techniques.

  • Monitoring: Continuous monitoring of AI systems.

  • Audit: Regular audits of AI models and processes.

 

PreviousNext