By João Moura, CEO and Co-Founder, Fraudio
In many payments businesses, fraud and AML still operate as two separate worlds. Fraud teams are built for speed, managing real-time authorisation rates, chargeback spikes, and point-of-sale risk. Compliance teams are built for scrutiny, working through transaction monitoring, investigations, retrospective reviews, and regulatory reporting.
For decades, the financial sector treated this division as a logical separation of duties. Fraud was categorised as a direct loss-prevention problem that impacted company profits, while money laundering was handled as a legal obligation managed through routine compliance checks. That separation worked when payment risk moved more slowly. In today’s environment, defined by instant settlement, global payment flows, and highly organised crime, it has become a serious liability. A siloed approach does more than slow the business down; it creates the conditions for fraudulent and suspicious activity to go unnoticed.
A Clash of Timelines
The core issue with this traditional split is a fundamental mismatch in operational speed. Modern payments increasingly move in real time. Transactions can now clear and settle in seconds, while many AML and compliance processes are still designed around retrospective reviews, batch analysis, and post-event investigation.
This shift alters the mechanics of risk management. When funds transfer instantly, after-the-fact analysis becomes less useful as a first line of defence. If compliance workflows rely on reviewing day-end records, the system transitions from a preventative shield into a ledger of what went wrong. By the time a traditional compliance tool flags a suspicious pattern, the funds may already have been dispersed across multiple accounts, platforms, or jurisdictions.
However, addressing this timeline gap does not mean turning AML into a real-time kill switch. Fraud prevention and compliance serve fundamentally different operational mandates. Fraud teams must intervene instantly to block and decline illicit transactions at the gate. AML teams, by contrast, are required to identify, assess, monitor, and manage risk over time. Abruptly blocking a payment under an AML flag risks alerting the customer, potentially compromising ongoing regulatory or law enforcement investigations.
The objective, therefore, is not to shift AML toward real-time blocking, but to equip it with real-time context. While the final actions remain distinct—fraud teams decline transactions, while compliance teams monitor, escalate, or file Suspicious Activity Reports (SARs)—the intelligence they rely on must be seamlessly connected. Compliance can no longer operate purely as a delayed, offline filter; it needs live signals from the transaction environment to build an accurate, continuous picture of risk.
The Operational Reality of Legacy Rules
The need for shared context becomes clear when looking at the daily reality of compliance operations. Historically, transaction monitoring has relied heavily on rigid, static thresholds, such as triggering an automatic review for any transaction crossing a specific monetary baseline. While these rules ensure regulatory compliance, they lack nuance. A legitimate, high-value transfer—like a high-net-worth individual purchasing luxury goods or a homeowner paying for a renovation—looks identical to a high-risk anomaly.
As a result, compliance departments spend significant resources investigating false positives, which in many organisations account for well over 90% of alerts. Sifting through these repetitive warnings creates a heavy operational burden, slowing down genuine investigations and diluting focus. Regulators want institutions to demonstrate effective risk management, but legacy frameworks built on siloed, static data often create operational inefficiencies that make genuine threats harder to identify.
How Criminals Exploit the Gaps
While financial institutions remain confined by internal silos, organised criminal networks face no such structural drag. Modern financial crime operates as a highly coordinated, automated industry. These networks understand corporate risk structures intimately, deliberately designing their attacks to slip between the cracks that separate fraud teams from compliance.
A clear example of this is the rise of fraud where transactions clear standard security checks undetected; either because the perpetrator is using a legitimately onboarded account, or because they have manipulated a genuine user into authorising a transfer. To an isolated fraud detection system examining immediate transaction variables, the activity appears entirely normal, and the payment is approved without suspicion.
The malicious nature of the event only becomes apparent when evaluating broader behavioural patterns over time, the exact perspective traditionally managed by compliance. It might manifest as a sudden change in cash-flow velocity, a strange correlation within a network of seemingly unrelated accounts, or a sequence of cross-border micro-payments.
Because the front-end tool lacks long-term history, and the AML system cannot see the live transaction stream, the pattern remains invisible until the loss is realised. This is exactly the type of gap sophisticated financial crime is designed to exploit. Criminal networks do not think in terms of fraud, AML, merchant risk or compliance. They simply look for the weakest point in the payment chain.
Why More Software Isn’t the Answer
The industry’s traditional response to emerging threats has been the continuous accumulation of software. Whenever a new risk appears, organisations tend to acquire an isolated tool to address that specific vulnerability. This has led to a fragmented patchwork of applications, leaving teams to manage separate vendors for transaction risk, merchant monitoring, and identity verification.
This approach creates a self-defeating cycle. The problem is rarely a lack of technology; it is a lack of shared context between the tools already in place. Amassing disconnected solutions does not deliver better risk management. It simply generates a high volume of contradictory alerts and unorganised data. When models are trained on narrow, restricted datasets and static parameters, false-positive rates inflate and internal review teams become overwhelmed.
Resolving this friction requires shifting focus away from individual software patches and toward core infrastructure. Risk mitigation cannot exist as a collection of separate, add-on tools. Instead of relying on simple, isolated rules, it needs to be built around a connected intelligence layer that understands relationships between transactions, merchants, accounts, devices, behaviours, and historical patterns in real time.
Turning Risk Management into a Growth Tool
When you remove the line between fraud and compliance data at the infrastructure level, the entire nature of risk management changes. It stops being viewed as a defensive expense or a compliance burden that slows operations, and instead becomes a powerful engine for growth.
In a highly competitive payments market, speed and conversion are paramount. If an acquirer or payment platform relies on heavy, fragmented controls, they introduce barriers for legitimate users; delaying merchant onboarding, triggering unnecessary hurdles at checkout, and blocking good transactions out of an excess of caution.
A single, connected platform provides that complete view of transaction risk. By analysing behavioural patterns and mapping connections across the entire payment chain in real time, the system gives fraud teams the precision to block genuine threats instantly, while giving compliance teams the rich, uncorrupted context they need to monitor and investigate without tipping off users or drowning in manual alerts. This gives businesses the confidence to safely optimise their authorisation rates, accelerate merchant onboarding, and scale into new markets without exposing themselves to unexpected regulatory penalties or sudden financial losses.
That is the commercial opportunity behind convergence. Better risk intelligence does not only help payment businesses stop more bad activity. It helps them say yes to more good activity with confidence.
The financial sector has reached a point where the separation of fraud and AML data is no longer sustainable. The threats we face are fast, connected, and constantly moving, and our security must match that reality. By breaking down these internal silos and treating risk management as a single, central infrastructure, companies can finally stay ahead of sophisticated threats and clear a safe path for sustainable growth.
Image provided by Fraudio
