• Principles-led approach: Stakeholders support a flexible, technology-neutral framework to manage AI risks, avoiding rigid, prescriptive rules.
• Integration with existing frameworks: AI-related risks should be addressed within current governance, risk management, operational resilience structures and third-party oversight requirements, rather than creating standalone AI-specific regulations.
• Proportionality: Governance measures should align with the risk profile of AI use cases, balancing innovation with risk mitigation.
• Bermuda’s market context: Regulatory approaches will reflect Bermuda’s predominantly institutional and cross-border market, avoiding retail-focussed models.
• Governance and accountability: Boards remain accountable for AI outcomes, with an emphasis on proportionate AI literacy and oversight mechanisms.
• Market integrity: AI’s use in trading and market surveillance raises concerns about systemic risks, requiring potential sector-specific guidance or supervisory clarification for higher-risk use cases.
• International alignment: The BMA will align with global regulatory standards to support firms operating across jurisdictions.
• Third-party risks: The Authority is considering whether existing third-party risk management frameworks sufficiently address AI-related dependencies on external providers, particularly regarding transparency, auditability, and vendor concentration.
• Implementation challenges: Phased adoption and supervisory engagement will support effective AI governance, particularly for smaller institutions.

The BMA will monitor AI adoption within existing frameworks, engage with stakeholders, and assess the need for incremental regulatory enhancements. The goal is to foster responsible innovation while safeguarding financial stability and market integrity.

For further details, refer to the full discussion paper here.