Day 6 - The Hard Truths Part 4: Why We’re Losing Ground Against Financial Crime
Overwhelming False Positives
Today, we’re talking about a massive blind spot in our efforts to fight financial crime: false positives. Imagine trying to find a single needle in a haystack, only to have someone keep adding more hay. This is what our compliance and law enforcement teams face every day. Detection tools, initially designed to help identify criminal behavior, have become so overactive that they’re generating mountains of low-quality alerts, drowning teams in work and burying real threats.
False positives are more than just an inconvenience—they’re an urgent issue draining resources, eroding morale, and weakening our capacity to address real threats. If we want to reclaim the integrity of our Anti-Financial Crime (AFC) systems, we must address the staggering cost of these inaccuracies and reimagine detection.
The Reality of False Positives: Why Our Systems Are Clogged with Noise
The concept of an alert system is simple: red flags should be raised when something unusual or risky happens. Yet, as detection tools have proliferated, so has the noise. In a typical AFC setup, detection systems rely on rule-based models that trigger alerts based on set criteria. But as crime patterns evolve, these systems struggle to keep up, often casting a wide net to avoid missing any potentially suspicious behavior. The result? A flood of alerts, the vast majority of which don’t indicate actual crime.
The numbers paint a daunting picture. According to a report by PwC, financial institutions discard 95-98% of generated alerts as false positives. That’s like spending all day shouting “fire!” in a crowded theater, with only a fraction of those shouts indicating an actual blaze. This is not just an ineffective system—it’s a dangerous one. When resources are consumed by false positives, real threats slip through unnoticed, making it harder to protect our financial systems and the people who rely on them.
How Did We Get Here?
False positives have been a part of AFC from the start, but their prevalence has grown alongside the industry’s focus on compliance. As regulatory requirements multiplied, detection systems became increasingly conservative, favoring overreporting to avoid penalties or oversight criticism. In an environment where non-compliance comes with heavy fines, the logical choice for many institutions has been to err on the side of caution, flagging anything that might raise concern. But what started as caution has turned into a crippling overload.
Think of it as a security guard who begins stopping every single person who enters a building because one might be a threat. Eventually, this over-caution becomes counterproductive, creating so much friction that genuine security threats are harder to detect. By prioritizing quantity over quality, AFC efforts have drifted from their original intention—identifying real criminal behavior—into a landscape where compliance has become synonymous with excessive screening.
The Costs of Excessive False Positives
The price we’re paying for this approach is far greater than most realize. Let’s break down the true impact:
1. Wasted Resources and Escalating Costs
2. Operational Inefficiencies and Increased Friction
3. Missed Opportunities to Prevent Real Crime
Why False Positives Have Become Tolerated
How did we allow false positives to become such an accepted part of AFC efforts? Much of it comes down to a culture of caution and a fear of penalties. When regulators assess an institution’s compliance performance, they rarely critique for overreporting . In fact, the industry has been conditioned to see overreporting (Defensive SARs) as preferable, or at least safer, than the risk of missing a reportable incident. Perhaps that's why Thomson Reuters is reporting record SAR filings .https://www.thomsonreuters.com/en-us/posts/investigation-fraud-and-risk/special-report-suspicious-activity-reports/
Furthermore, outdated detection systems are hard to replace. Many FIs still rely on legacy rule-based systems designed years ago, with criteria that may no longer be relevant but continue to run because of inertia or cost barriers to updating them. These systems are too rigid to adapt to evolving crime tactics, yet switching to more agile systems requires significant investment. As a result, many institutions tolerate high levels of false positives because they see no viable alternative.
What We Need to Do Now: Redefining Detection with Precision
If we are to reclaim AFC from the swamp of false positives, we need to make serious, foundational changes. Here’s how we can shift from noise to precision:
1. Invest in AI-Driven and Machine Learning Solutions
2. Collaborate with Tech Innovators and Regulators to Set Standards
3. Shift the Focus to Quality, Not Quantity
4. Foster a Proactive, Rather Than Reactive, AFC Approach
Moving Forward: Breaking Free from the Noise
False positives are more than an inconvenience—they’re a roadblock in our mission to protect the financial system from crime. If we continue to tolerate this issue, we risk losing sight of our purpose, overburdening our teams, and allowing real threats to thrive.
The time has come to address the reality of false positives with the urgency it demands. This series is focused on identifying and confronting the roadblocks in our current approach, and false positives are among the biggest barriers we face. But there is hope. With the right investments, collaborations, and mindset shifts, we can reduce noise and refocus on what matters: real crime prevention.
Stay tuned as we continue to dig into the underlying issues holding us back from effective AFC and begin mapping out actionable solutions. Together, we can move from a world of overwhelming alerts to one of precision, clarity, and real protection.
Tomorrow: Day 7 - The Hard Truths Part 5 - Lack of Focus on Victims and the Impact
Compliance & Fraud Expert | BSA/AML/CFT/OFAC Specialist | AVP Financial Crimes Risk Mgmt | Training & Consulting
1 周I want to copy/paste my comment from Day 3. Great read again. I look forward to Day 7 while appreciating days 1-6.
Founder, Global Regulatory Financial Crime Compliance, Financial Crime Technology, Machine Learning and GenAI, Blockchain/Crypto AML, CyberFraud across Banking, Payments, Insurance, Asset Management and Wealth.
1 周Richard Stocks - CAMS thanks again for another day. I also will add my opinion is AI systems in place for detection are hiding true negatives and that is even worse. But because of the opaquenesss they are generating false positives that we don't even yet know or can audit. I also don't believe these solutions will have accuracy as I believe the whole approach to AFC is based on poor and siloed data signals. The vendor claims of AI are let down by poor implementation and data constraints at financial institution. I agree that a lot transformation and rethinking is needed before even modern tech truly helps in the fight against financial crime.