A lack of infrastructure, challenging geography and financial exclusion from the formal economy have created barriers to traditional financial services models in Africa. Innovation and digital transformation have driven the rapid expansion of financial services across the continent, especially over mobile devices by telecommunications providers, bringing transactional power to customers in the most rural areas. 

However, digital disruption to meet changing customer needs has also opened up new avenues for financial criminals to exploit. The event shed light on the tools and platforms that allow financial services organisations to strike the optimal balance between meeting their growth objectives and reducing risk.

Total transaction volumes using digital channels in Nigeria more than doubled from 2018 to 2020, from 1.3bn transactions to 3.3bn transactions. According to the Nigerian government, digital payment channels helped to support continued business activities during Covid-19 lockdowns and continue to evolve to meet the needs of both businesses and households. 

However, nearly 36% of adult Nigerians still do not have access to financial services to conduct digital payments, access credit and ease operating conditions for small businesses. The country’s Central Bank’s deployment of its first digital currency, the E-Naira, reflects the government’s commitment to providing a reliable channel for remittance flows and cross-border payments.

“There is no doubt that the digitalisation of financial services will boost financial inclusion and benefit Africa’s people, but many competitors face challenges in identifying and combating fraud. The problem is that it is impossible for humans to proactively monitor transactional activity and trends to tackle financial crime. No organisation has enough human resource capacity,” says Babalola Oladokun, Regional Lead for SAS in Nigeria.

The answer, Oladokun says, has to be deployment of artificial intelligence and machine learning. “Globally, since the onset of Covid-19, we have seen a spike in sophisticated financial crime, including social engineering, malicious use of AI, identity theft, payment fraud and SIM swapping. Data and analytics are the only viable tools to combat this surge, by providing financial institutions with automated algorithms that incorporate a cross-channel view of customer behaviour and geolocation of users to spot complex fraud trends.”

Dynamic behavioural profiles and adaptive machine learning combine to ensure that organisations can stay up to date with changing fraud tactics and control them in real time instead of learning about them the hard way. 

Stephan Wessels, SAS Head of Customer Advisory for South Africa, says financial institutions generate a wealth of data every day, which can be harnessed to fight fraud and understand customer behaviour better. “The benefit of being able to scale the business quickly is, unfortunately, the fraudster’s avenue for exploitation. Lockdown restrictions and working from home have suddenly exposed organisations and employees to technologically advanced fraudsters who are also scaling their own operations using cloud-based technologies.”

According to Wessels, money laundered annually is estimated at 2%-5% of global GDP, equating to several million dollars per minute. “The banking industry’s effectiveness rate at detecting true suspicious activities remains under 5%. This means many institutions are suffering terrible financial and reputational damage from not providing greater protection to customers. Automation is the only way to improve this level of accuracy.”

Controls such as stronger authentication can be incorporated into digital channels without disrupting the customer experience, says Oladokun. “In conjunction with AI flagging abnormal transactions that differ significantly from user patterns, have not taken place at a place where the person is known to be, or are conducted through a new device, this provides the institution with the power to stop transactions within milliseconds or decline payments. It also reduces the number of false positive alerts, which significantly reduces the error rate and frees up investigative resources to focus on risky activity.”

 Machine learning methods that include deep learning neural networks are proving to be more accurate and effective than rules-based approaches.

 “AI provides the capability to learn from complex data patterns and interactions to create holistic views of customer activity that continue to strengthen defences. Analytically driven business decisions are better decisions. The real clincher for financial services institutions is that analytics can help them understand their customers better, bringing benefits beyond fraud prevention,” concludes Wessels.