Fraud prevention is often discussed in terms of technology: better rules, faster alerts, stronger authentication and more sophisticated monitoring.
But what happens after fraud actually occurs?
How should an institution decide which incidents deserve the greatest attention? When should law enforcement become involved? How should compliance teams respond when a customer genuinely authorised a payment but was manipulated by a fraudster? And how does an institution ensure that every fraud incident makes its controls stronger rather than simply becoming another closed case?
These were some of the questions explored at Regfyl’s inaugural When Fraud Happens roundtable in October 2025.
The closed-door session brought together fraud and compliance professionals from across Nigeria’s financial ecosystem to discuss the operational realities of fraud management.
The conversation was held under the Chatham House Rule, allowing participants to share experiences, institutional challenges and practical lessons that are rarely discussed publicly.
What emerged was a clear message: effective fraud management requires much more than detecting a suspicious transaction. It depends on how institutions classify fraud, prioritise incidents, design products, engage law enforcement, learn from losses and collaborate with the wider ecosystem.
1. Institutions Need a Common Language for Different Types of Fraud
One of the starting points for the discussion was that not all fraud happens in the same way.
A customer who is manipulated into sending money to a fraudster presents a fundamentally different problem from a criminal who opens an account using a synthetic identity. Both situations are also different from an account takeover in which a legitimate customer loses control of their account.
Treating all of these simply as “fraud” can make it harder to determine what controls are needed and who should be responsible for responding.
The roundtable introduced Regfyl’s AFIU Fraud Framework, which provides a practical way to distinguish between four broad categories of fraud.
A — Authorised Fraud
Authorised fraud occurs where the genuine customer authorises the transaction but does so because they have been deceived or manipulated.
Examples include business email compromise, romance scams, investment scams and other forms of social engineering.
The institution may have correctly authenticated the customer and still fail to prevent the fraud because the problem is not identity theft. It is manipulation.
F — First-Party Fraud
First-party fraud occurs where the customer themselves commits or facilitates the fraud.
This could involve deliberately false applications, fraudulent claims, misuse of products or situations where the customer knowingly assists a wider fraud scheme.
The challenge here is very different from protecting an innocent customer. The institution is trying to detect dishonesty by the person it has onboarded.
I — Identity Fraud
Identity fraud involves the use of false, stolen or synthetic identities.
The weakness may originate during onboarding, where falsified or compromised information allows a fraudulent account to enter the institution.
This type of fraud reinforces the importance of identity assurance, not simply document collection.
U — Unauthorised Fraud
Unauthorised fraud occurs where a third party gains access to an account, payment channel or device without the genuine customer’s permission.
Examples include phishing, SIM swaps, stolen credentials and account takeover.
The controls relevant here may include authentication, device intelligence, behavioural monitoring and account-security measures.
The value of the AFIU framework is not simply classification.
It helps institutions recognise that different fraud types require different responses.
Better authentication may reduce account takeover but do little against a customer who willingly sends money after being deceived. Transaction monitoring may identify unusual behaviour but cannot fully compensate for weak onboarding that allowed a false identity into the institution.
Effective fraud prevention therefore begins with understanding what type of fraud the institution is actually dealing with.
2. Fraud Prioritisation Should Go Beyond the Amount Lost
Fraud teams naturally focus on monetary value.
Large losses attract attention because the financial impact is immediately visible.
But participants agreed that value alone is an incomplete measure of fraud severity.
A relatively small incident may expose a significant control weakness, a new fraud typology, insider involvement or a vulnerability that could affect thousands of customers.
It may also create regulatory, reputational or customer-trust consequences that far exceed the original financial loss.
The more useful question is therefore not simply:
“How much money did we lose?”
It is:
“What does this incident tell us about our exposure?”
A ₦500,000 fraud that reveals a weakness capable of being replicated across a large customer base may ultimately be more important than a one-off ₦20 million incident with no wider systemic implication.
This is why fraud prioritisation needs to be risk-based.
Institutions should consider the financial impact, but also the control failure, customer harm, regulatory exposure, reputational consequences and likelihood of recurrence.
3. Fraud Response Does Not End When the Transaction Is Stopped
Fraud response is often framed around the immediate objective.
Can the transaction be stopped? Can the account be restricted? Can the money be recovered?
Those questions matter, but participants stressed that the work often continues long after the initial fraud event.
Once an incident occurs, the institution may need to coordinate with law enforcement, preserve evidence, produce internal case documentation, respond to regulators, manage customer complaints, investigate possible staff involvement and prepare for litigation.
There may also be a broader control review.
Did the fraud expose a weakness in onboarding? Was the transaction visible to monitoring systems? Were relevant alerts investigated quickly enough? Did staff follow internal procedures? Are other customers exposed to the same vulnerability?
Preparation for these downstream consequences is therefore just as important as the first response.
An institution that acts quickly in the first hour but has no structured process for the days and weeks that follow may still struggle to manage the incident effectively.
4. Authorised Fraud Requires a Different Approach
Authorised fraud was one of the most important issues discussed during the roundtable.
This occurs where the legitimate customer makes the transaction themselves but has been manipulated into doing so.
A fraudster may convince the customer that they are speaking with their bank, that an investment opportunity is genuine, that a supplier’s account details have changed or that money must urgently be moved to keep it safe.
From a traditional payment-security perspective, the transaction may appear completely legitimate.
The real customer authenticated.
The real customer initiated the payment.
The real customer approved it.
Yet the transaction is still the product of fraud.
That creates a very different control problem from account takeover or credential theft.
Participants discussed whether Nigeria may eventually move towards stronger frameworks around Authorised Push Payment fraud and reimbursement, while recognising that any such framework would need to deal carefully with questions of liability, evidence and customer responsibility.
The broader lesson, however, is already clear.
Authentication alone is not enough to establish that a transaction is safe.
Institutions increasingly need to understand context.
Is this a new beneficiary? Is the amount unusual for the customer? Has the customer recently changed device? Is the transaction inconsistent with their historical behaviour? Is the receiving account connected to known fraud activity?
This is where behavioural intelligence becomes increasingly important.
5. Fraud Prevention Should Start at Product Design
Many fraud controls are introduced too late.
The product is designed, the customer journey is built, commercial teams launch it and only afterwards do fraud and compliance teams begin identifying vulnerabilities.
Participants argued that the process should work differently.
Fraud prevention should begin at the product-design stage.
When risk and compliance teams are involved early, institutions can build appropriate guardrails into the customer journey rather than trying to patch vulnerabilities after launch.
Depending on the product, those controls could include stronger verification for certain transactions, additional checks for higher-risk customers, transaction limits, cooling-off periods, behavioural monitoring, device intelligence or stronger escalation requirements.
The principle is simple:
Risk awareness should be designed in, not patched in.
This can reduce both fraud losses and the cost of remediation later.
6. Every Fraud Incident Should Improve the Control Environment
A fraud case should not simply move from:
Open → Investigated → Closed
and then disappear.
Every incident contains intelligence. What allowed the fraud to happen? Which controls worked? Which controls failed? Was the activity visible before the loss occurred? Did transaction monitoring produce an alert? Was the customer’s identity properly established? Did staff follow internal procedures? Could another customer be exposed to the same weakness?
Participants discussed the importance of capturing these lessons systematically rather than relying on institutional memory.
A strong feedback loop should look like:
Incident → Investigation → Root Cause → Lesson → Control Improvement → Monitoring
Over time, this creates institutional resilience.
If the same fraud succeeds repeatedly through the same vulnerability, the organisation is resolving individual cases rather than actually improving its fraud framework.
7. Law Enforcement Engagement Should Begin Before a Crisis
Institutions often engage law enforcement only after a serious fraud has occurred.
The roundtable highlighted the value of more proactive engagement.
Regular interaction with agencies and financial-intelligence authorities can help institutions understand evidential requirements, investigation processes, information requests and practical expectations before a major incident occurs.
It can also help institutions establish more effective internal procedures.
When a law-enforcement request arrives, the organisation should already know who receives it, how authenticity is confirmed, which internal teams need to be involved, who approves disclosure, what documentation must be preserved and how the interaction is recorded.
The middle of a fraud crisis is not the ideal time to design those procedures for the first time.
Proactive engagement also helps build mutual understanding between financial institutions and law-enforcement agencies about what information is available, what evidence is useful and how quickly institutions can realistically respond.
8. Fraud Prevention Cannot Succeed in Institutional Silos
Fraudsters rarely operate within the boundaries of one financial institution.
Funds may leave one bank, enter another institution, move into a fintech wallet, pass through a payment processor and then be split across several additional accounts.
Each institution may see only one part of the activity.
The fraudster sees the entire chain.
That creates an information asymmetry that individual institutions cannot fully solve alone.
Participants repeatedly returned to the importance of collaboration over competition.
The discussion recognised the potential value of shared fraud databases and industry intelligence initiatives, while also acknowledging the practical challenges around privacy, data quality, governance, timeliness and integration.
The aim is not unrestricted sharing of customer information.
It is to create lawful and effective ways for institutions to share relevant fraud indicators and emerging typologies quickly enough to reduce the time available for criminals to exploit the same method across multiple institutions.
The principle is straightforward:
An institution fighting ecosystem-wide fraud using only its own internal data will always have an incomplete picture.
9. Institutions Need a Defined Fraud Risk Appetite
Fraud decisions can easily become inconsistent when every incident is handled independently.
One team may escalate a loss that another team would consider routine. One manager may prioritise financial value while another focuses on reputational impact.
Institutions therefore need a clear fraud-risk appetite.
This means defining the kinds of financial loss, customer harm, regulatory exposure, reputational damage or systemic weakness that require escalation.
Clear thresholds help create more consistent decision-making and provide management with a defensible basis for allocating resources.
Fraud-risk appetite also has a direct relationship with budgeting.
Effective fraud prevention requires investment in technology, analytics, training, investigations, customer education and external engagement.
Without intentional funding, even highly capable fraud teams can remain trapped in reactive case management.
A fraud strategy without an appropriate budget is difficult to operationalise.
The Bigger Lesson: Fraud Management Is an Institutional Capability
The strongest theme emerging from the first When Fraud Happens roundtable was that fraud cannot be treated as an isolated technical issue.
Effective fraud management touches:
Product → Technology → Compliance → Risk → Legal → Operations → Customer Service → Law Enforcement → Executive Management
The institution needs not only tools capable of detecting suspicious activity, but also governance that determines what happens next.
It needs to learn from incidents.
It needs to collaborate externally.
And it needs to make fraud risk part of product and business decisions before losses occur.
Download the Full When Fraud Happens Insights
The full When Fraud Happens: Insights and Updates document captures additional observations from the October 2025 closed-door roundtable, including the AFIU Fraud Framework, authorised fraud, law-enforcement engagement, product design, fraud-risk appetite and industry collaboration.
Download the When Fraud Happens Insights
The conversation did not end with the first roundtable.
Regfyl subsequently brought industry participants together again for When Fraud Happens II, where the discussion shifted from how institutions respond to fraud towards the emerging threats and structural challenges shaping fraud prevention in Nigeria.