Optimize Your Tech Stack Without Disrupting AML Performance
How AI Transforms AML/CFT
Executive Summary
Yaron Hazan
Vice President of Regulatory Affairs
Fintechs and banks today share more similarities than differences. Both operate in fast‐moving digital ecosystems, process increasingly complex financial flows, and face rising expectations from regulators, partners, and customers. While fintechs have driven innovation through speed and flexibility, banks bring depth, scale, and decades of compliance maturity. Yet both sectors now confront the same reality: financial crime has evolved, regulatory scrutiny has intensified, and legacy compliance frameworks, whether within a bank or a high‐growth fintech, are struggling to keep pace.
Today, regulators across the US, UK, EU, Nordics, LATAM, and emerging markets have converged on the same expectation: that both must operate AML programs with the same rigor, explainability, and governance — regardless of size or growth stage.
Legacy detection systems and manual investigative workflows are struggling under the weight of rising alert volumes, increasingly complex typologies, and global expectations for transparency. The consequence is an operational model that is costly, inefficient, and reactive — creating drag on growth and exposing institutions to enforcement risk.
But modernization no longer requires a multi-year infrastructure overhaul. AI enables institutions to strengthen detection performance, reduce false positives, and scale efficiently — without replacing core platforms.
This white paper explores:
- Global regulatory escalation and why AML now defines institutional risk
- The operational drag cycle and its impact on cost, customer experience, and risk exposure
- How AI reshapes detection and decisioning without system replacement
- Regional requirements (US, UK, EU, Nordics, CEE, LATAM) and how institutions can align
- How to optimize existing systems while preparing for future regulatory demands
- A real-world transformation case study
- ThetaRay’s practical roadmap for modernizing compliance without disruption
In an environment where financial crime adapts quickly and regulators expect transparency, institutions must adopt technologies and operating models that deliver both stronger risk coverage and scalable efficiency without compromising agility or customer experience.
ThetaRay enables this by helping customers continuously optimize their detection programs; reducing noise and refining scenarios so teams can focus on the highest-value risks. Our optimization approach ensures institutions maintain robust coverage while achieving measurable operational efficiency and faster, more accurate decision-making.
Foreword
As financial institutions enter 2026 and beyond, compliance and financial crime teams face a clear mandate: strengthen risk coverage while operating with greater efficiency and resilience. Expectations are converging across jurisdictions around risk‑based monitoring, robust governance, and now more than ever, auditability. At the same time, many AML programs are under growing operational strain. Rising alert volumes, manual investigation processes, and fragmented workflows can create compounding inefficiencies that inflate cost and delay timely detection and risk response.
Fortunately, institutions can modernize AML programs without undertaking disruptive, multi year system replacements. Across traditional financial institutions, the most practical path forward is emerging as optimization rather than overhaul.
By applying advanced AI as an enhancement layer on top of existing monitoring frameworks, institutions can improve detection quality, enhance risk coverage, and scale operations without disrupting core platforms or introducing unnecessary risk. This optimization first approach preserves operational continuity and maintains clear lines of accountability, which is critical for regulators and auditors assessing how new technologies are introduced and governed.
Sustainable improvements in AML effectiveness require more than the introduction of new technology. Meaningful performance gains depend on how well AI is integrated into existing operating environments, including data quality, case management processes, analyst workflows, model governance, and change management. Without this alignment, even advanced technologies may fail to deliver a measurable or lasting impact. This is where the right advisory partnership becomes essential, bridging regulatory expectations, operational realities, and technological capabilities to deliver outcomes that are validated, auditable, and scalable.
In this white paper, ThetaRay outlines an optimization first roadmap that enables institutions to enhance performance without disruption, validate improvements through parallel testing, and scale AI driven detection with confidence.
Attribution:
Alex Methot, Manager, Matrix USA
Amit Kabra, Director, Matrix USA
Table of contents
Across the world, regulatory expectations for financial institutions, whether traditional banks, fintechs, or payment companies, are rapidly converging. Supervisory bodies are aligning around a common set of principles: transparency, accountability, explainability, and risk-based decision-making. This shift is driven by the dual pressures of increasing financial crime complexity and the accelerating speed of money movement services.
Regulators in major jurisdictions are now harmonizing expectations around model governance, detection quality, risk-based monitoring, and operational resilience. Cross-border information sharing is increasing, and enforcement actions demonstrate that fast growth without strong controls results in severe financial, operational, and reputational consequences.
In today’s landscape, any institution facilitating money movement must meet the same AML standards, regardless of size, licensing category, or business model.
Common expectations emerging across the U.S., UK, EU, Nordics, and CEE include:
-
1
Robust risk-based monitoring aligned to FATF principles
-
2
Explainability for AI-driven decisions and risk scoring
-
3
Strong governance frameworks with clear auditability and human oversight
-
4
Operational resilience and continuity during technology transition
-
5
Accurate, timely SAR/STR filing
-
6
Operational resilience and continuity during technology transition
-
7
Consistent detection quality across products, geographies, and customer types
Frameworks such as SR 11-7 (U.S.), FCA PS24/17 (UK), the AMLR and EU AI Act (EU), and instant payment fraud rules (Nordics) explicitly mandate explainability, transparency, and governance for advanced analytics. The EU AI Act and UK’s FCA emphasise model documentation and justification, while U.S. bodies such as FinCEN, FFIEC, and the Federal Reserve highlight intelligence-led, risk-based programs supported by mature oversight. Meanwhile, Nordic and CEE regulators increasingly demand real-time detection capabilities to mitigate risks in instant-payment and high-growth ecosystems.
As a result, institutions operating across multiple jurisdictions can no longer rely on fragmented, region-specific compliance strategies. Instead, they must adopt globally consistent, explainable, and adaptable detection systems that satisfy multiple supervisory environments simultaneously.
Regional Snapshots: How Expectations Are Shifting
United States
U.S. regulators have significantly escalated oversight for both banks and fintechs, with a focus on TM maturity, SAR quality, governance, scalability, and third-party risk
Key supervisory bodies: FinCEN, CFPB, OCC, Federal Reserve, and state regulators (NYDFS, DFPI, Texas).
Recent enforcement actions (Cash App, Wise US, OKX) highlight rising expectations for:
- Scalable, risk-based monitoring
- Faster and higher-quality SAR decisioning
- Mature governance and documentation
- Consistent detection across all customer segments
Banking industry: Traditional banks face mounting pressure to modernize legacy systems and justify models under SR 11-7.
Fintech industry: Fintechs—and their bank partners—must demonstrate equivalent standards and governance maturity.
United Kingdom
The UK is asserting global leadership in financial crime and responsible AI regulation.
Key frameworks: FCA PS24/17 (explainability & auditability), SMCR (senior accountability), BoE/FCA AI guidance, and payments/open banking oversight.
Banking industry: Banks must demonstrate continuous oversight and explainability for AI-driven models.
Fintech industry: Fintechs must provide bank-grade transparency and documentation to maintain sponsor bank trust and permissions.
European Union
The EU is undergoing its most sweeping AML transformation since AMLD4
AMLA : centralized supervision across the EU
AMLR : unified rulebook applied directly to all institutions
EU AI Act : AML models classified as “high risk,” requiring explainability and human oversight
Banking industry : Large banks face intensified, harmonized cross-border scrutiny.
Fintech industry : Fintechs must meet identical standards to retain corridor access and passporting rights.
Nordics
With some of the world’s most digitized financial ecosystems, Nordic regulators prioritize:
- Real-time detection in instant-payment environments
- Proven risk based approach for AML
- High data integrity and resilience
- Transparency and model explainability
Banking industry: Large banks remain under stringent supervision following historic AML failures.
Fintech industry: Neo-banks and PSPs must prove operational resilience and comparable oversight.
Central & Eastern Europe (CEE)
CEE financial systems are scaling rapidly while tightening controls. Key drivers include:
- Strengthened VASP regulation (Lithuania, Estonia, Czechia)
- Increased cross-border flow complexity
- Heightened scrutiny of digital identity and onboarding systems
- More frequent inspections for high-growth PSPs
Banking industry: Local banks are upgrading systems to retain correspondent relationships.
Fintech industry: Payment companies must demonstrate scalable AML operations to avoid de-risking.
What This Means for Institutions
To meet the new regulatory reality, financial institutions must adopt technologies and operating models that deliver stronger risk coverage, explainability, and scalable efficiency, without sacrificing agility or customer experience. ThetaRay’s Cognitive AI enables institutions to increase accuracy, reduce false positives, and maintain robust governance, helping them stay ahead of evolving regulatory expectations across all jurisdictions.
Regulatory alignment (Global comparison table)
Table: What regulators now expect
Clear.Bank’s Journey to 10x growth and efficient financial crime compliance
Clear.Bank—a tech-bank focused on creating the best and most sustainable banking and payment infrastructure in the world—operates with the mandate to build trust by combating financial crime effectively and efficiently. Their commitment to innovation and excellence has placed them ahead of the curve in a rapidly evolving financial landscape.
Background and Challenges
Clear.Bank recognized that traditional compliance approaches generated excessive noise and consumed significant resources. With plans to scale rapidly, anticipating a tenfold increase in transaction volumes within a year, the bank needed a solution that could grow with its business while preserving strong compliance standards.
“ThetaRay is certainly enabling our growth. In the year’s time, we’ll be doing 10x the volumes that we’re doing today.”
Justin Pascoal-Reynolds
Program Director Clear.Bank
ThetaRay’s Cognitive AI Solution
ThetaRay’s Cognitive AI Transaction Monitoring solution now provides Clear.Bank with enhanced risk coverage by detecting suspicious behaviors across customers, accounts, and transactions spanning multiple business lines. This precision enables the bank to focus resources on real risks without reducing oversight.
The platform significantly cuts operational noise by reducing false positives and lowering the volume of alerts needing manual review. Investigations are faster, more accurate, and analysts are more productive. Advanced network insights also improve risk visibility, revealing relationships and patterns that previously required extensive time to uncover.
With ThetaRay, insights that once took weeks now surface in minutes, reducing investigative workload while maintaining high effectiveness.
Operational scalability
Clear.Bank is well-positioned to handle a significant increase in transaction volumes without compromising on compliance or efficiency.
Trust and credibility
Enhanced compliance measures have reinforced customer confidence in Clear.Bank’s commitment to fighting financial crime.
Cost savings
By reducing false positives and investigator workload, Clear.Bank achieved significant cost efficiencies while maintaining robust risk management practices.
Strategic growth enablement
ThetaRay’s Transaction Monitoring has empowered Clear.Bank to pursue its global expansion with confidence.
“Financial crime compliance is a balancing act between risk coverage and doing it effectively and efficiently. And I can comfortably say that ThetaRay allows us to do that.”
Yasemin Swanson
Chief Operating Officer Clear.Bank
Clear.Bank’s experience demonstrates how advanced Cognitive AI technology enhances compliance, drives operational efficiency, and supports sustainable growth in a rapidly evolving environment.
Leading financial institutions are increasingly adopting a phased transition to AI, one that balances innovation, regulatory confidence, and operational continuity.
Why Optimization Comes First
Legacy transaction monitoring engines remain deeply embedded across payments, KYC, sanctions screening, and case management workflows. Fully replacing them introduces material operational and regulatory risk, including data migrations, extended validation cycles, parallel testing, and retraining across jurisdictions.
As a result, optimization, not immediate replacement, is emerging as the most practical first step.
AI as an Optimization Layer: Easier for Regulators and Auditors to Digest
Applying AI on top of existing controls is significantly easier for regulators and auditors to assess.
Once risks are defined by subject-matter experts, typologies documented, and controls implemented, AI functions as an enhancement layer, not a replacement of regulatory intent. This allows auditors to clearly map requirements to controls, while institutions improve effectiveness without introducing compliance risk.
As Yaron Herzan, VP Regulatory Affairs at ThetaRay, explains:
“From a regulatory perspective, optimization is often the most responsible way to introduce AI. When institutions apply AI on top of existing, well-understood controls, regulators can clearly see continuity, governance, and accountability, while still benefiting from better detection and efficiency.”
Optimization is a strong first step in the transition from rules-based transaction monitoring to AI-based monitoring.
Industry Direction: Expected Over the Next Decade
Based on industry estimates and analyst outlooks, transaction monitoring is expected to undergo a structural shift over the next decade:
Today:
Approximately three out of four financial institutions still rely heavily on a rules-first approach to transaction monitoring
Looking to the decade:
- An estimated ~75% of transaction monitoring is expected to be AI-driven
- Total cost of ownership for AML programs is expected to decline by up to ~50%, driven by improved detection accuracy, reduced false positives, and greater automation
These figures reflect directional industry expectations, informed by regulatory guidance, market analyses, and the growing pressure on institutions to improve effectiveness while controlling compliance costs.
ThetaRay’s Cognitive AI Optimization Approach
ThetaRay’s approach aligns directly with this industry trajectory by enabling institutions to modernize without disrupting existing operations.
Our Cognitive AI overlays existing monitoring environments and is fundamentally different from rules or supervised machine learning:
These figures reflect directional industry expectations, informed by regulatory guidance, market analyses, and
the growing pressure on institutions to improve effectiveness while controlling compliance costs.
ThetaRay’s Three-Stage Optimization Model
1
Optimize & Enhance Existing Compliance Operations
Start by enriching your current systems with AI. Overlay Cognitive AI on incumbent rule engines to cut false positives, reduce noise, and improve investigative outcomes, all without disrupting workflows. Analysts benefit from prioritised, explainable alerts that boost efficiency and shorten investigation time.
2
Parallel Testing & AI-Driven Optimization
Run AI models in parallel with legacy rules to formally validate performance. This challenger approach allows institutions to gradually minimise rules to obvious risks while regulators see transparent, explainable results. AI-driven optimisation improves accuracy and builds confidence in transitioning away from static logic.
3
AI-Led Detection
Once trust and performance are established, Cognitive AI becomes the core detection engine. Legacy systems are decommissioned or limited to simple scenarios, reducing long-term costs and technical debt. The result is a future-proof compliance architecture that scales with business growth.
A Strategy Aligned with Regulators
This model isn’t just technological, it’s strategic. It lets institutions move confidently from rules-based monitoring to AI-driven detection at a pace that preserves operational continuity, reduces cost, and elevates AML effectiveness.
The global financial ecosystem is undergoing rapid transformation, with regulators worldwide intensifying scrutiny and financial crime becoming more dynamic, at the same time, the “defense” mechanism is left outdated, old and paralyzed.
Institutions must modernise compliance in a way that balances innovation with stability. ThetaRay’s Cognitive AI solution offers a path to modernization. By enhancing existing systems at the first stage, significantly increasing efficiency and opening the way to modern detection. Institutions that modernize today, position themselves not only to meet current expectations but to lead in an industry where trust, intelligence, and agility define long-term success.
In short, the future of compliance isn’t rule based, optimization is an excellent first step transforming AML from rules to AI based.
