Consultation Paper - Guidance Note - The Responsible Use of Artificial Intelligence in Bermuda's Financial Services Sector (2026-08-14)

Summary

This is a Bermuda Monetary Authority (BMA) Consultation Paper setting out a draft Guidance Note on the responsible use of Artificial Intelligence (AI) by entities the BMA regulates. It follows a July 2025 Discussion Paper and stakeholder feedback, and asks the market to comment on a proposed principles based, proportionate approach to AI governance before the Guidance Note is finalised.

The draft Guidance Note does not create new statutory or regulatory obligations or a separate AI approval regime. Instead it explains how existing legislative, regulatory and supervisory requirements (governance, conduct, cyber risk, operational resilience, outsourcing, AML/ATF and sanctions frameworks) apply when a regulated entity develops, procures, deploys or otherwise uses AI, whether built in house, obtained from a third party or group affiliate, or embedded in another product or platform.

• Five proposed outcomes: Accountable governance and risk ownership; proportionate lifecycle management; reliable, explainable and appropriately overseen AI outcomes; secure and resilient deployment (including third-party arrangements); and adequate evidence for internal assurance and supervisory review.
• Key subject areas covered: Strategic direction and risk appetite, governance roles, AI inventory and documentation, materiality and risk assessment, model/system selection, data governance, validation and testing, change management, human oversight, explainability and fairness, generative and agentic AI, cybersecurity, operational resilience, third-party/group risk, conduct and disclosure, AML/ATF and sanctions, recordkeeping and supervisory engagement.
• Proportionality: Expectations scale with the materiality and risk of the specific AI use case and the size, complexity and resources of the entity, rather than a one-size-fits-all AI taxonomy.
• Consultation ask: BMA seeks feedback on whether the guidance provides sufficient clarity, whether the proportionality framework is operationally feasible, and whether further sector-specific or emerging risks need to be addressed before finalisation.

As a consultation paper in draft status, the Guidance Note is not yet in force. Compliance officers should review the proposed text now and consider submitting comments, since the final Guidance Note will shape how existing obligations are interpreted for AI use cases once adopted.

Key obligations

• Interested parties wishing to respond to the consultation must submit comments to policy@bma.bm by 30 October 2026.
• Once in force, regulated entities using AI would remain responsible for complying with their existing applicable legal and regulatory obligations regardless of whether the AI system is developed internally, procured from a third party, or provided by a group affiliate.
• Under the proposed framework, entities would be expected to maintain governance, risk ownership, AI inventories/documentation, human oversight, and recordkeeping proportionate to the materiality and risk of each AI use case, to support internal assurance and supervisory review.

Applies to

insurance and reinsurance entities, captives, insurance-linked securities entities, investment funds, digital asset business entities, trust companies, corporate service providers, fund administrators, banking and money service businesses, other entities regulated and supervised by the BMA

Deadlines

• 30 October 2026: Deadline for submitting consultation responses on the draft Guidance Note to policy@bma.bm.