Webinar Description
Key Takeaways
- Examines why traditional annual compliance reviews are becoming insufficient for AI-driven processes
- Analyses regulatory signals from the Bank of England, European Systemic Risk Board and European Central Bank
- Addresses the governance challenges posed by agentic AI systems that make autonomous decisions
- Compares diverging UK and EU approaches to AI oversight in financial services
- Designed for compliance officers, risk managers and governance professionals in regulated industries
Introduction
The webinar “When the Regulator Moves Faster Than the Law” addresses a fundamental shift in how financial regulators expect organisations to demonstrate compliance. As agentic AI systems—those capable of making autonomous decisions within regulated processes—become more prevalent, supervisory bodies are moving away from accepting periodic compliance snapshots. This 45-minute session explores what continuous oversight means in practice and how compliance frameworks must evolve to meet these expectations.
The timing reflects a broader industry challenge. Regulators including the Bank of England, the European Systemic Risk Board and the European Central Bank have signalled that static, point-in-time evidence is no longer adequate for processes where AI behaviour can change rapidly between review cycles. For compliance and risk professionals, understanding these signals is essential to avoiding regulatory friction.
About This Event
This online webinar is led by Tomas Hellum, Head of Regulatory Strategy at Decision Focus. The session takes a practical, checklist-driven approach to help attendees assess their current compliance posture and identify gaps in their oversight frameworks. Rather than theoretical discussion, the focus is on actionable steps organisations can take immediately.
The Shift from Periodic Reviews to Continuous Evidence
Traditional compliance programmes in financial services have relied heavily on annual or quarterly reviews. An organisation would document its controls, conduct testing at defined intervals, and present evidence to auditors and regulators on a scheduled basis. This model assumed that processes remained relatively stable between assessments.
Agentic AI disrupts this assumption. These systems can adapt their behaviour based on new data, potentially altering risk profiles in ways that would not surface until the next scheduled review. Regulators have responded by expecting what the webinar describes as “dynamic, continuously generated compliance signals”—real-time or near-real-time evidence that controls remain effective as AI systems operate.
This transition requires more than technology upgrades. It demands rethinking how compliance teams structure their monitoring activities, how they document evidence, and how they communicate with supervisors who increasingly expect to see live dashboards rather than retrospective reports.
Human-on-the-Loop Oversight for AI Systems
A central theme of the webinar is the concept of human-on-the-loop oversight. Unlike human-in-the-loop models where people approve each decision, human-on-the-loop approaches allow AI systems to operate autonomously while humans monitor outputs and intervene when necessary. This distinction matters for compliance because it changes how organisations must document accountability and escalation procedures.
New AI risk classifications are emerging that require organisations to categorise their AI applications by autonomy level and potential impact. Higher-risk classifications trigger more stringent oversight requirements, including more frequent monitoring and clearer escalation paths when systems behave unexpectedly.
Navigating Divergent UK and EU Requirements
Organisations operating across jurisdictions face the additional complexity of diverging regulatory approaches. The UK and EU have taken different paths on AI governance, creating what the webinar frames as “two clocks, one control set”—the challenge of satisfying multiple regulatory timelines and expectations with a unified compliance framework.
For multinational financial institutions, this divergence creates practical difficulties. Building separate compliance programmes for each jurisdiction is costly and inefficient, yet a single framework must be flexible enough to accommodate different supervisory expectations. The session addresses strategies for managing this complexity without duplicating effort.
Who Should Attend
The webinar is designed for professionals responsible for governance, risk and compliance in regulated industries. This includes compliance officers, risk managers, regulatory strategists, heads of audit, and IT and AI governance professionals. Senior executives in financial services, insurance, banking and fintech who need to understand the strategic implications of these regulatory shifts will also find the content relevant.
Attendees from mid-to-large organisations are likely to benefit most, as these entities typically face the greatest scrutiny from supervisors and have the most complex AI deployments to govern.
Conclusion
As regulatory expectations outpace the formal rulemaking process, compliance professionals must anticipate supervisory direction rather than wait for codified requirements. This webinar offers a framework for understanding where AI governance is heading and practical guidance for organisations seeking to stay ahead of enforcement trends.