Why Transforming Sanctions Screening is Critical for Banks

As global conflicts escalate and regulatory watchlists expand by the day, banks are being squeezed between rising enforcement pressure and the explosive growth of real-time transactions. One misstep can trigger multi-million-dollar fines or lasting reputational damage. In this environment, traditional screening systems simply can’t keep up.

The cost of inaction is rising – and the time for Agentic AI has arrived!

A discipline of increasing complexity

Screening for sanctions is a vital element of a bank’s compliance strategy. By identifying and blocking transactions and customers involving sanctioned or high-risk individuals, entities, countries and sectors, banks greatly reduce their risk of getting into regulatory hot water, and protect their reputation.

But the sanctions landscape is becoming increasingly complex and changing fast. Regulatory requirements and watchlists are constantly evolving while bad actors continue to develop sophisticated evasion tactics. At the same time, the volume of sanctions is rising, with the European Union alone issuing 679 financial sanctions last year.[1]

Increasing volumes, and decreasing processing times for payments and onboarding, now mandate more automated compliance checks to detect and prevent financial crime. Our clients must avoid fines and protect their reputation from unintended violations. Financial crime compliance is a strategic shield for regulatory and reputational trust.

It’s no wonder that many banks are struggling to stay on top of it all, particularly when there are significant costs involved.

Rising costs of sanctions screening compliance

According to Celent’s 2024 Report on IT and Operational Spending on Financial Crime Compliance, global financial institutions spent an estimated $34.7 billion on financial crime compliance in 2024, with sanctions accounting for about 20% of this total.[2] And when you factor in the time spent investigating false positives, costs are even higher.

It’s well documented that the financial (not to mention reputational) consequences of non-compliance are also steep. US banks faced over $33 million in sanctions-related fines and settlements in 2023 – over 18 times the total in 2022. Between 2021 and September 2023, individual banks paid an average of over $4 million in sanctions violations.[3]

Now is the time for change

On top of the costs of compliance (and the even greater cost of non-compliance), several factors are complicating the screening process.

For one, the rapid growth and demands of instant payments mean banks have a much narrower window in which to process and validate transactions. They are under pressure to screen a growing volume of transactions at speed while maintaining a high degree of accuracy.

Geopolitical events and global conflicts have also caused a surge in sanctions as governments increasingly use them as foreign policy tools. This means watchlists are expanding and changing all the time. Regulators have upped their oversight with stricter enforcement and stiffer penalties for non-compliance, but perhaps most importantly is their intensifying focus on explainability and auditability. Banks need to be able to show regulators not only how their screening processes work, but how individual decisions were made and why – especially when AI is used to help decision making.

Although human oversight will always be essential for high-risk or complex investigations, manual workflows are proving too slow and prone to error for today’s sanctions volumes. This is contributing to the pervasive issue of false positives – alerts that unnecessarily create extra work for compliance teams, delay legitimate transactions, and, not least, frustrate customers. For many banks, outdated screening systems are compounding the problem. Often inflexible and reliant on manual intervention, they can push the false positive rate far beyond the industry average of 5-8%.[4]

Smarter solutions for sanctions screening

With mounting pressure on resources and increasing demands for quicker turnarounds, banks need reliable, scalable tools to help manage sanctions screening and the associated risks.

When embedded into watchlist screening, Agentic AI goes beyond passive data processing or single task automation. Goal directed, it can autonomously make reasoned decisions, probabilistically modelled, which are explainable, learn, and adapt:

• Reducing false positives by matching names and entities more accurately across global and domestic watchlists;
• Automating alert triage, routing low-risk cases for automated clearance and escalating high-risk alerts for human review
• Strengthening compliance with systems that are explainable, auditable, and aligned with regulatory standards.

AI’s role in sanctions screening

There is no denying that AI is playing a growing role in banking compliance. Research suggests that, today, some 60% of banks use AI for risk management and fraud detection.[5] And according to the Bank of England, AML and fraud detection are seen as the second-most impactful AI use cases over the next three years.[6]

Still, AI is not a silver bullet. Human expertise remains essential to interpret insights, navigate (sometimes very nuanced) compliance requirements, and make well-informed decisions that can be justified in front of a regulator. The true value comes from empowering teams with tools that enhance their ability to respond quickly, confidently, and accurately to information and risk.

Temenos FCM AI Agent is designed to help you achieve this. With the industry average false positive rate sitting at around 6%, our solution, combining the precision of AI with human oversight, could bring this rate down to 1-2%.

By automating watchlist screening, it not only lowers false positives; it also helps to optimize compliance workflows and process more transactions in real time without delays or unnecessary investigations.

Resources:

[1]https://data.europa.eu/apps/eusanctionstracker/

[2] https://www.celent.com/en/insights/445011014

[3] https://complyadvantage.com/insights/sanctions-screening-banking/

[4] https://financialcrimeacademy.org/sanctions-screening/

[5] Modernizing Banking Solutions, Hanover Research for Temenos, April 2025

[6] https://www.bankofengland.co.uk/report/2024/artificial-intelligence-in-uk-financial-services-2024