By Elizabeth Callan, SymphonyAI Financial Crime & Compliance SME, North America
The risk-based approach is broken. Agentic AI is the fix.
For over two decades, financial institutions have been told to adopt a risk-based approach to AML compliance. In practice, most programs remain static, rules-driven, and backward-looking. This means that they are risk-based in name only. This whitepaper makes the case for a fundamental redesign: replacing periodic assessments and manual threat monitoring with continuous, intelligence-driven risk orchestration powered by agentic AI.
The financial crime threat landscape doesn't pause between annual reviews. This whitepaper explores how agentic AI and large language models enable institutions to close the gap between emerging threats and operational response for good.
Key takeaways include:
The question is no longer "What is the risk level for this customer or product?" It's "What is the evolving risk posture — and how are our controls responding right now?"
This whitepaper is authored by Elizabeth Callan, a recognized leader in the fight against money laundering and financial crime with over 25 years of experience across intelligence, law enforcement, and the private sector.
At SymphonyAI, Elizabeth drives strategy and innovation, developing AI-led, intelligence-driven solutions that are transforming how global institutions detect, disrupt, and prevent money laundering, sanctions evasion, and sophisticated financial crimes. Her work is helping shift the industry approach from reactive, technical compliance to proactive, risk-aligned, and highly effective risk management.
Prior to SymphonyAI, she served as a Senior Intelligence Analyst at the U.S. Department of the Treasury, advising senior officials at OFAC and FinCEN, and as Intelligence Liaison and Senior Advisor to the DEA's Special Operations Division. She holds a master's degree in economics and the ACAMS Advanced Certification in Financial Crimes Investigations (CAMS-FCI).
Note: Elizabeth has also recorded an on-demand webinar on this topic - Re-engineering the risk-based approach in AML compliance
In this paper, you'll learn how to:
The era of the periodic risk assessment is ending. Unlock the blueprint for continuous risk alignment, where agentic AI and human expertise work together to keep your institution ahead of financial crime, not catching up to it.
Symphony Risk Intelligence
Introducing Symphony Risk Intelligence - From Reactive to Proactive Risk Management
Reinventing the Compliance Operating Model
Command and Control Rewired: Agentic AI in Anti-Financial Crime
Guide to Explainable AI in Financial Services
The AI-native FinCrime platform designed to help financial institutions move from reactive to proactive risk management.
A continuous risk alignment approach replaces the annual or biannual risk assessment cycle with an always-on intelligence function that dynamically maps emerging threats to an institution's specific risk profile. Rather than asking "what was our risk last year?", compliance teams can understand how their exposure is shifting in near real time and respond accordingly.
Unlike rules-based automation, agentic AI can read and interpret unstructured information — such as regulatory advisories, enforcement actions, and typologies — and reason about what those threats mean for a specific institution's products, customers, and geographies. It then translates that intelligence into structured, actionable risk insights without requiring constant human prompting.
When designed with the right guardrails, agentic AI can actually strengthen regulatory defensibility by improving the consistency, documentation, and traceability of risk decisions. Key governance principles - including human-in-the-loop oversight, explainable outputs, clear accountability, and model validation - are essential design requirements, not optional add-ons.
FATF, FinCEN, FINTRAC, the FCA, and the Wolfsberg Group have all issued guidance in recent years calling for institutions to move beyond static, procedural compliance toward intelligence-led, outcome-focused risk management. The direction of travel is consistent globally with regulators wanting to see programs that demonstrably work in practice, not just on paper.
The whitepaper outlines a practical framework — the Continuous Risk Assessment-Control Calibration Cycle — that breaks the transformation into manageable stages, from threat intelligence ingestion through to autonomous control recalibration. SymphonyAI's team of global financial crime experts are also available to discuss how an incremental adoption path using Symphony Risk Intelligence could work for your institution specifically.
Elizabeth has spent more than 20 years tackling money laundering (ML) and financial crime. At SymphonyAI she drives the strategy and innovation that delivers transformational compliance solutions. Prior to SymphonyAI she worked within the U.S. intelligence and law enforcement communities. As a Senior Intelligence Analyst with the U.S. Department of the Treasury, she drove U.S. policy and enforcement actions and supported U.S. officials and policymakers, including at OFAC and FinCEN, on ML threats and sanctions initiatives. She also served as Treasury’s first Intelligence Liaison and Senior Advisor to DEA’s Special Operations Division, spearheading large-scale ML investigations and intelligence collection initiatives, training law enforcement agents and analysts, and promoting collaboration between Treasury and U.S. and foreign law enforcement. In the private sector, Elizabeth also worked within financial institutions and consulting managing investigations teams, developing risk management strategies for complex products and services, and designing institutional AML programs and controls. Elizabeth also teaches AML and sanctions courses at the university level.
.svg)