Dr Mark Nasila - Afrikan Heroes

How AI Could Transform Financial Services in Emerging Markets

August 20, 2022

By Mark Nasila

Financial inclusion is a global challenge. Solving it has enormous benefits, not only for the financial institutions that can offer new customers access to products and credit they may previously not have had, but for empowering the unbanked, enabling them to access financial services, and assisting them in financial planning while offering them hitherto inaccessible means to pursue their aspirations.

A society with greater financial inclusion is also good for the wider economy. McKinsey estimates that using digital financial services for inclusion alone could add US$3.7-trillion to an emerging economy’s GDP in less than a decade.

Dr Mark Nasila, is chief analytics officer in First National Bank’s chief risk office — **Dr Mark Nasila, chief analytics officer in First National Bank’s chief risk office**

The World Bank says that an inclusive financial system is fundamental when it comes to reducing extreme poverty, promoting sustainable inclusive economic growth and development, and boosting shared economic prosperity. Why? Because greater financial inclusivity allows those previously excluded to borrow, save, invest and start businesses, while also positioning themselves to weather socioeconomic disruption.

Generally speaking, to secure a line of credit, a person needs a financial history, and a detailed one at that

More than two billion people globally do not use formal financial services, in part because financial services providers haven’t found the right way to approach them, or been sufficiently motivated to. This is because the unbanked are deemed expensive to provide services to, and it’s assumed there is little prospect of them eventually providing the sort of upside to make them worth the effort.

ResearchGate found that despite improvements to African banking services in recent years, the continent’s ratio of private credit to GDP remains far behind that of other markets. It averaged 24% of GDP in sub-Saharan Africa in 2010 and 39% in North Africa, which is far lower than the 77% average of all other developing economies, and the 172% average of developed ones.

The problem is no better if one zooms in on individual access to finance. A paltry 23% of adults in Africa have a bank account. But it’s worth noting this figure varies considerably between countries and regions. For instance, in South Africa, the figure is 51%, while it plummets to 5% in the Central African Republic.

Fewer than a third of adults globally have access to any sort of credit bureau, even though the World Bank estimates that two-thirds of the global unbanked nonetheless have access to a mobile device. This is key because lack of credit history isn’t only a problem for adults; it’s one millennials looking to buy cars or houses, or fund education, also face. But new approaches can help alleviate this problem. For instance, machine learning and big data can be used by new fintech players (or established ones) to create credit scores for consumers who don’t have sufficient data points in the traditional financial system.

Alternative data

Generally speaking, to secure a line of credit, a person needs a financial history, and a detailed one at that — simply opening a current or savings account is seldom sufficient. In those places where children or teenagers get accounts opened for them by their parents, and where those same parents are willing and able to act as guarantors, access to credit is taken for granted. But in cases where the parents themselves have no conventional banking products, credit is all but impossible to come by.

Artificial intelligence can change this by allowing financial services providers to make use of alternative data sources to assess creditworthiness. For individuals, meaningful inferences can be made by looking at data on their smartphone (with their permission) which can reveal financial behaviour, from call activity to app usage, and even which applications a consumer uses most. In agriculture, meanwhile, satellite images can be used to estimate past and future income from farming and make decisions about loans accordingly.

This opens the door to new avenues of lending: for example, smartphone-based microlending, bereft of the usually prohibitive and punitive interest rates such lending models tend to use to insulate themselves against risk. Because AI can construct pseudo networks of similar people, and combine complex network analysis with representation learning, it can also make assessments with ethical concerns and privacy regulations in mind, while avoiding the sort of discrimination that often dogs human-based decision making.

The ever-increasing affordability and adoption of smartphones opens up entirely new market segments to financial services providers.

For instance, the time and location of transactions — and other spatiotemporal traits — have a surprisingly wide range of predictive value. Similarly, the social networks (or social graphs) of individuals can provide impressively accurate and actionable signals of credit quality.

The ever-increasing affordability and adoption of smartphones opens up entirely new market segments to financial services providers. The increasing use of social media provides another avenue for alternative data because the connections between people can reveal all sorts of probabilities. For example, a person’s most contacted friends can influence the likelihood of them abandoning a service if those friends have previously done so themselves.

Alternative algorithms

In conjunction with alternative data, alternative algorithms can improve risk assessment when dealing with consumers with limited financial data to offer. Logistic regression is already commonplace in banking, but by adding additional vectors like random forests, assessment models can account for features previous models might discount or miss and be better equipped to identify those that are statistically relevant.

For example, a California-based start-up called Tala uses 250 data points, including consumers’ online behaviour, cell phone and data usage, and other metrics, to assess their desirability as clients. The process begins with identity verification: is the prospective customer who they say they are? Thereafter, it uses supplied data to make a decision on a customer’s loan request.

Tala customers who are approved for loans are incentivised to pay them off in less than the allotted period, something microlenders often penalise customers for, and personal data is deleted after the loan is repaid. Instead of getting users to resubmit the data they did originally for any subsequent loans, their previous repayment behaviour is used for assessment.

Similarly, a US service called Wemimo uses utility bill and rent payments to help create credit scores for the estimated 45 million Americans without a traditional credit history, the lack of which generally excludes US citizens from getting loans for cars, houses, or other high-cost purchases or expenses.

This also opens the door for lenders to enter into partnerships with utility providers, service providers, wholesalers, telecommunications providers, or even governments to identify promising data sources for assessing potential customers’ creditworthiness (with those customers’ explicit consent). Alternatively, lenders can help these repositories of user data to become lenders themselves, something many retailers already do by using a loyalty card first to assess spending behaviour and then offering an adjacent credit service to those customers it identifies as suitable candidates.

Another player in this emerging space is Branch As, a credit product that uses mobile payment system M-Pesa as its delivery mechanism. Like Tala, Branch As requires prospective customers to access data on their smartphones via a mobile application they download. AI makes automatic decisions and adjusts interest rates (between 2% and 14% monthly) accordingly. Newcomers to the service receive smaller loans with higher fees, but subsequent loans adjust depending on users’ repayment behaviour, and all successful loans are deposited directly into the users’ M-Pesa account, so they can be available in seconds.

Open APIs (application programming interfaces) from telecoms services providers have made these sorts of credit assessments easier and more reliable by providing third parties with phone usage and bill payment information. In Nigeria, for example, using telco data provides access to credit to whole swathes of the population to whom it would otherwise be out of reach. While fewer than 3% of Nigerians are eligible for conventional bank loans, nearly 50% of them have mobile phones, suggesting a credit scoring system predicated on mobile data isn’t just feasible, it has the power to be massively empowering.

One of the alternative algorithms that make this sort of assessment reliable and useful is reinforcement learning. By using new information to adjust thresholds, reinforcement learning can not only make more accurate decisions on creditworthiness but can also proactively guide customers to optimal behaviour so as to improve their odds of a successful application. Equally, it can be used to flag, preempt and respond to consumers’ financial distress or other risks.

Studies by M Herasymovych et al (November 2019) found that dynamic reinforcement learning systems — that is, those which constantly adapt their credit-decisioning thresholds in response to live data feedback from customers — tend to reduce the sorts of biased results that can lead to eligible customers being misclassified and denied credit erroneously.

Handling imbalanced datasets

A scarcity of data or credit history in some population groups might lead to datasets being imbalanced. If one class is smaller than another, for example, the larger of the two may be favoured unduly. If a model is weighted to non-defaulters, for instance, defaulters may be unreasonably biased against. Imbalance datasets make assessment more difficult and bias more likely to creep in.

Various techniques have emerged to help to mitigate this problem. These techniques range from random-under-sampling and random over-sampling to Synthetic Minority Oversampling Technique (Smote) that tries to provide more accurate filler data by randomly creating data points that are near but not identical to the source ones. Another solution is changing the performance metrics. These can include using a confusion matrix, adjusting for precision or recall, or using a variety of algorithms and comparing their results rather than relying on a single one.

But there aren’t the only means of injecting fairness into machine-learning (ML) models. Researchers at MIT have used a new technique that reduces bias in image-recognition systems, even when ML models are trained on unbalanced datasets. The solution, called Partial Attribute Decorrelation (Parade) trains the model to evaluate certain metrics separately.

Parade can be applied to other metrics and can similarly flag sensitive ones so that they don’t become reductive measures by which images, creditworthiness or risk are assessed.

Digitising access

Digitising access and creditworthiness, especially in the micro-finance sector, can drive financial inclusion significantly. Aside from reducing bias, it can also improve access, because while many people may struggle to physically get to a branch, most have access to digital channels and platforms.

The solution may lie in a blended or hybrid approach, where some decisioning and loan functions are driven by AI, but others still employ human judgment and interactions. This is particularly relevant where access to digital channels doesn’t translate into comfort using them, and may mean the difference between achieving a competitive advantage and squandering one.

Another company using phone-based assessment is CredoLab, a Singapore-based fintech company which does scoring on lenders’ behalf using smartphone metadata, developing scorecards it can apply to similar applications. CredoLab’s solution has resulted in a 20% higher rate of customer approvals of new clients, a 15% reduction in non-performing loans.

NLP applications in finance

Natural language processing (NLP) is a technology that allows humans and machines to communicate using the sort of language encountered in everyday life. It shows up in everything from smarthome assistants to chatbots and even interactive voice recordings. But it’s also finding a growing number of use cases in finance.

NLP can also be used to assess not only creditworthiness, but trustworthiness, something that’s traditionally hard for humans to do. It can do this by executing deeper analysis on interactions, and by understanding hidden patterns in speech or language that may be too subtle for humans to detect, but which can be recognisable from huge sample sizes gleaned from previous, similar interactions. This approach has been successfully deployed by Microbnk and Capital Float, among others.

In addition to risk assessment, NLP can be used in chatbot advisors that can offer customers credit advice and coaching, or which can assist customers to understand products or specific details about them. NLP also makes it possible to provide services in multiple languages, which in turn can improve consumers’ comprehension and comfort — something that’s especially valuable where complex financial services are involved, like home loans.

Bank of America’s Erica chatbot, for example, is available via the company’s mobile application to more than 25 million customers. Erica acts not only as a financial assistant — reminding customers to make credit card payments or pay bills — but also acts as a financial advisor in conjunction with another product called Life Plan, which focuses on helping BoA customers achieve their financial goals, from purchasing a home to paying for their children’s education or minimising the impact of their existing debt.

AI can do far more than improve financial inclusions. Once consumers are within the financial system, it can help them make the most of the products and services on offer to them. At the same time, it can help financial services providers to reach new customers, provide them with the right products and services, all while reducing risk.

Dr Mark Nasila, is chief analytics officer in First National Bank’s chief risk office

Kelechi Deca

Kelechi Deca has over two decades of media experience, he has traveled to over 77 countries reporting on multilateral development institutions, international business, trade, travels, culture, and diplomacy. He is also a petrol head with in-depth knowledge of automobiles and the auto industry

How to disrupt with data in the technology era By MARK NASILA

August 18, 2020

From punch cards that logged workers’ hours in early factories to the basic computers that powered the space race, it’s evident that we’ve been living in a data-driven world for more than a century. But changes in technology in the last two decades mean we now need to use data in more inventive and innovative ways if we want to be genuinely disruptive.

The era of emerging technologies is changing the way we live, work and connect with and relate to one another. At the same time, businesses face changing expectations from their customers, and governments have to contend with new demands from their citizenry.

***Dr Mark Nasila is chief analytics officer in FNB’s chief risk office***

We are at an inflection point where emerging technologies are on the cusp of becoming mainstream, and the borders between physical, digital and biological technologies are becoming harder to delineate. Neurotechnology, artificial intelligence and robotics have begun to bleed into one another. So, too, have 3D printing, bioprinting and rapid prototyping.

Neurotechnology, artificial intelligence and robotics have begun to bleed into one another. So, too, have 3D printing, bioprinting and rapid prototyping

In a little more than half a century, we’ve moved from computers that occupied huge rooms to cloud computing and artificial intelligence. At the same time, the scope, quantity and quality of data being generated by the systems embedded in our lives are unprecedented. The annual rate of data generation, meanwhile, will grow from 16 zettabytes (or a trillion gigabytes) this year to 160 zettabytes per year in 2025, and of that data, more than a quarter of it will be created in real time, with 95% generated by Internet of things devices.

Human data generation is going to grow exponentially in parallel to machine-generated data, and ubiquitous connectivity is going to become more than expected; it’s going to be demanded. Because of this, there’s an expectation from customers around what they expect companies to offer them. One-size-fits-all solutions will no longer be sufficient.

Harness data to offer bespoke solutions

For consumers, exclusivity now lies in customisation, and in some industries, like luxury fashion, in scarcity. The growth of personal data about customers also means sectors that wish to be disruptive need to successfully harness that data, make sense of it, and deliver desirable products or services on the back of it.

Some of the industries successfully using data to drive disruption include security, surveillance and defence (China’s mass surveillance systems, military drone tech and airport security), medicine, biology, and agriculture (seeding, drone assessments, and doctors performing or guiding surgery remotely), and education and training (augmented and virtual reality as teaching tools).

Others also benefiting include autonomous vehicles (Tesla owners can summon their vehicles via an app, or use its screen to watch Netflix while it’s parked), and building and construction (architecture and design, AI-generated floor plans, or digital quantity surveying). The lesson in all of these is that, if you want to create a different future, you have to use data in a different way.

Disruption versus innovation

A culture of disruption is more important than one of innovation. Innovation alone is iterative. It expands on existing solutions, rather than totally upending or reinventing them as disruption does. Innovation, by and large, makes existing things better. Disruption makes entirely new things.

When people talk about “disruption”, they often actually mean innovation that derives from first principles thinking. That is, the sort of innovation that stems from going back to the drawing board. Jeff Bezos incorporated first principles into his business model, urging entrepreneurs to “resist proxies” like established processes or market research and instead strive toward more fundamental truths, like real outcomes or customer needs.

Steve Jobs, meanwhile, got his designers and engineers to “think different” (sic) and rework Apple’s systems from scratch. The result wasn’t a phone; it was the entire Smartphone category.

SpaceX’s Falcon Heavy:Elon Musk perfected reusability and slashed costs

Another example is SpaceX’s Falcon Heavy rocket compared to the classic Saturn V used by Nasa. The assumption with the Saturn and its kin was that the rockets used for space launches necessarily had to be disposable. SpaceX CEO Elon Musk thought that reusable rockets might be possible. He’s been proven correct, and drastically reduced the cost of getting payloads into orbit as a result.

One of the key differentiators that allowed SpaceX to undercut conventional solutions is its use of 3D printing in engine manufacture. Reducing the cost of fabrication while simultaneously increasing the lifespan of its hardware means SpaceX is able to execute launches for 10% of the cost of the Apollo space programme, this could be the difference between merely visiting space and inhabiting other planets.

SpaceX’s ability to achieve Apollo-like results for cutting costs by 90% is the result of the culture of disruption Musk has fostered. The Apollo programme could only make incremental cost reductions through design by analogy. A breakthrough of SpaceX’s sort required fundamentally rethinking the problem of moving objects from the Earth’s surface into orbit.

Iterative thinking (or “design by analogy”), is the other dominant model for innovation, though some — like Musk — might argue it barely constitutes innovation at all. Iterative thinking means applying new solutions to old problems. For example, in the US banking system, because cheques remain an acceptable means of transferring value between individuals, many banks now let customers take photos of cheques with their smartphones to remotely cash them.

The danger of using new solutions to do what you’ve always done is that, because the inputs are the same, your outputs will be similarly homogenous. That’s why more than 80% of companies fail at digital transformation: They see the shift to digital as evolutionary, when in fact it’s revolutionary.

Think globally, thrive locally

If you need further convincing that data is fundamentally altering the business landscape, consider the growth in unicorns (that is, companies with a valuation in excess of US$1-billion) over the last decade. In the early 2000s, there were a handful of unicorns minted each year. Between 2014 and 2017 there were almost 200.

Why does this matter? Because it means that customers are being faced with ever more choices, and in doing so, those choices have to be increasingly compelling if they’re going to attract users, and more importantly, retain them. It also changes what we define as customer centricity and what competition looks like. The nature of this unprecedented speed, scale and scope of disruption means companies can no longer confine their attention to competitors locally; they have to consider their rivals globally.

Netflix, in one fell swoop, doubled its footprint by simply offering its content in 100 new markets. Meanwhile, digitally savvy consumers understand the options available globally, so even if they aren’t in a market with a particular product or service, they expect their local equivalents to at least try to be competitive. Customers with the mobility to operate globally, meanwhile, will simply take their business to the most agile and able provider.

Strategy is driven by people and culture

So, how do you drive disruption? First, you need to inspire confidence in the people in your organisation to try new things. At the same time, new initiatives must be driven by competent people who understand the problem you’re trying to solve. The culture and competencies required for disruption are different than those required for innovation.

Under Satya Nadella’s leadership, Microsoft is tackling social challenges and making the company socially responsible. It’s important to realise that it’s no longer enough to provide a service: Society expects you to contribute to solving broader, systemic problems. It’s no longer good enough to design solutions that only serve the business’s ends. Nor is it sufficient to measure success by revenue KPIs. Instead, one has to consider social KPIs.

It used to be that data and tech were used to design products and services and offer them to customers using different channels. This “omnichannel strategy” meant not only fostering relationships with customers through various means, it meant maintaining them that way. But today, the expected approach is a ubiquitous strategy.

Instead of an omnichannel strategy focused on pushing particular products or services, in a ubiquitous strategy you’re sharing the entire solution so that other people can adopt it. One of the advantages is it means customers identify you with an overarching solution rather than with a specific product.

Innovation can come from openness

In 2014, Elon Musk made many of his electric vehicle company Tesla’s patents free to use. Musk’s argument for sharing his company’s intellectual property was that his incentive is to get the automotive industry to shift to greener initiatives as soon as possible. Sharing IP made it more likely the goal would be achieved, and sooner. Search giant Google open-sourced its TensorFlow AI engine because it believes the problems it’s trying to solve are so large it can’t reasonably expect to do so alone.

In both examples, what matters is that the way success is measured is different from the conventional metrics of revenue generation or profitability. By measuring success in new ways and opening problems to new prospective solvers, businesses can harness hitherto unforeseen opportunities.

The global goalposts have moved

The sustainable development goals (SDGs) — also known as the Global Goals — were adopted by the member states of the United Nations in 2015. Their core tenets are to alleviate poverty, provide solutions to the ecological devastation the Earth faces and bring peace to war-torn regions. To realise them, companies need to adjust their strategies to focus on society, where how much they improve it is a metric of success at least as important as profitability.

It’s helpful to think of SDGs as brand or value propositions. When you’re solving a society’s systemic problems, you’re not just a philanthropist, you’re a social entrepreneur. For example, leaders at Davos this year heard that achieving the Global Goals could open an estimated $12-trillion in market opportunities in four economic systems: food and agriculture, cities, energy and materials, and health and well-being. Taking advantage of these opportunities fully — that is, maximising benefits for business owners, shareholders and other stakeholders while simultaneously maximising social benefits — will require inclusive business practices. This can be achieved by the fourth sector model which includes blended consortiums, public private partnerships, civic or municipal enterprises, cross sector partnerships, hybrid organisations, and social enterprises.

SDGs are now useful criteria for measuring your company’s success. There are 17 unique SDGs, and while it’s unreasonable and unrealistic to try and solve for all of them at once, some companies are solving for multiple goals at once. Musk, for instance, is solving for sustainable energy and environmental preservation. Microsoft, meanwhile, under Satya Nadella’s leadership, is tackling social challenges and making Microsoft socially responsible.

Furthermore, with this mindset in place, the long-term value of a company is no longer a matter of earnings alone. Human capital, innovation, and societal impact are all important yardsticks, too, and long-term value needs to be weighed against talent, happiness, sustainability and other previously intangible assets, all of which can contribute to financial success.

Engage your customers, enhance your brand

Customers want to be identified for who they are and how they behave. Whether through facial recognition or deep-learning techniques and habit analysis, customers expect businesses to know not just who they are, but how they operate, and what they want. Moreover, they expect this level of personalisation without having their security compromised.

Looking at an ID document for manual verification doesn’t make the document safer. Instead, biometric data plus behavioural information can be used to link a consumer to their actions and ensure that anomalous activity is spotted promptly. But the onboarding process isn’t something unique to financial services. Every organisation, business, enterprise, and government agency, large or small, has to manage onboarding processes, and can benefit from more effective ones. Users want to be identified as individuals, not numbers, no matter the sector.

The same holds true for repeat interactions.

While there are fears about biometrics and their potential misuse, blockchain networks are being used to solve this problem. Biometrics and behavioural identity can help provide a standardised identity link for first parties, but also for partnering with other services and solutions providers, all the while providing customised but applicable offerings. A customer can have different account or product numbers, but not different biometrics. Tying them together can provide oversight and contextual awareness that makes for better customer interactions, or more effective upselling.

For example, biometrics and payment history can be used in a retail environment to help direct customers to aisles based on their historic purchases, or to remind them about a rewards programme at checkout. In the case of Amazon Go’s retail stores, to take the idea to further reaches, AI-powered cameras and user accounts allow for interaction-less shopping and checking out, with customers billed automatically for what they take with them from the store. Simultaneously, Amazon gets to build an ever more detailed profile of a customer, enabling it to better target promotions and realise more consistent conversion rates. While there are fears about biometrics and their potential misuse, blockchain networks are being used to solve this problem, protecting identity information while ensuring only legitimate actors can use it.

The cloud is an enabler, not a buzzword

For cross-platform data services to work, you need cloud-based solutions. Not only does the cloud make it easier for a wide range of businesses to leverage each other’s networks, it ensures greater security than is possible when each company has to assess the reliability or the validity of data in isolation, and in-house.

Previously, data science techniques have provided a snapshot of a customer at a specific point in time. While these have done well in predicting outcomes based on that moment and guiding strategies around the predictions, they have not been sufficient to provide strategic insight on what is driving those outcomes. Nor have they assisted businesses to adjust their strategies to obtain different, desirable outcomes as opposed to expected, empirical ones.

But today, because there is lots of data available, engagement changes and understanding moves from standalone snapshots to a cumulative, and potentially predictive, customer journey, where it’s possible to infer a future need from a current one, and pre-emptively respond down the line. Moreover, it’s possible to recognise which behaviours lead to which outcomes, and then actively encourage the correct behaviour from a customer to direct them to the desired outcome.

Identify the motivation, shape the behaviour

One such technique enabled by new technologies is reinforcement learning (RL). RL is a machine-learning approach that points to the suitable action to take to maximise the reward in a particular situation. It is employed to find the best possible data-driven behaviour or path a subject should take in a specific situation.

In traditional, supervised learning the training data has the answer key built into it, and that answer is the intended outcome, which shapes the course of the learning accordingly. With reinforcement learning, there is no specific answer. Instead, the reinforcement agent decides what to do to perform the given task, which means that in the absence of a training dataset, it is bound to learn from its experience.

RL provides a way organisations can link observed, data-driven behaviour to social needs in the future. For instance, a customer’s transactional or investment behaviour may indicate they’re focused on future education, housing, travel or health needs. A financial institution that recognises this may then be able to shape its products or services in such a way that the customer is incentivised to adjust their behaviour accordingly.

At the same time, this prescience might require different channels. A customer might be best reached via e-mail during office hours or via mobile after hours. They might be best served by a phone call for some queries, or a chatbot for others. But what’s important is that their place in what might previously have been considered a relationship cycle should now be thought of as a series of unique circumstances, each of which determines the channel of engagement.

It’s not about making the decisions around the state of play, but about influencing it. Instead of creating an app because a customer might use it in future, it’s about creating an app because a customer will use it in future because it’s the best channel for their needs and they’re incentivised to use it, which means they do.

Use doesn’t mean contentedness

Just because a customer uses a product doesn’t mean they like it. How they use a product, their comments, complaints or social media posts about it, or even their interactions with call centres — which can be analysed by AI that understands customer’s complaints or comments, called natural language understanding (NLU) — can all lead to a more telling sentiment analysis that better indicates perception and satisfaction. NLU, in this case, is providing insight into customer data. The data just happens to be voice data rather than digits on a spreadsheet.

This sort of approach can also result in personalised solutions to unique problems or solutions to problems that prove not to be unique and can then be offered to other consumers before they become dissatisfied. These sorts of approaches allow for extremely granular analysis of how customers perceive your products, where their pain points are, and how their perceptions may have changed over time or may do so as their relationship with, and expectations of, your business evolves.

At the same time, these sorts of adaptable, predictive models can help financial institutions and other risk-sensitive businesses better manage and assess risk. Traditionally, risk has been managed at a single point in time, but now the structure of the problem has changed, and financial institutions want to try to help customers avoid getting into a risk space in the first place.

As threats become more complex, it’s necessary for warning systems and pattern recognition to do likewise. Similarly, predictive data models and accompanying technology can be used to monitor financial crime risks and regulatory requirements, including tax avoidance, money laundering and other organised crime activities than can leave financial institutions subject to punitive measures from government or other regulatory bodies. Mitigating these risks and developing a reputation for being safe and immune from abuse can also keep fraudsters at bay and encourage new customers to take up your services.

As threats become more complex, it’s necessary for warning systems and pattern recognition to do likewise. Enhanced due-diligence AI tools create holistic views of customers, using varied data sources, and creating a multifaceted set of insights that can be used to better predict and counter risks like fraudulent activities. These AI tools don’t just do quality assurance and make recommendations but can be used for financial advice, onboarding, strategy analysis, or designing individual rewards programmes and other unique and specific offerings where practical, or necessary.

Built to purpose means built to last

While it might seem the obvious course to take is to collaborate with fintechs, start-ups, universities and other higher education institutions with experience building AI solutions, the risk is that by outsourcing the design and implementation you’ll forever be beholden to outside contractors for your AI needs. Also, any unique IP created along the way won’t belong to you, and your ability to respond to changing needs will be hamstrung.

Wherever possible, when employing AI it’s preferable to build solutions in-house, from the ground up. That means less procuring, more building, and far more opportunity to customise and add to your solution over time. The result is an AI model that matches your needs and your societal challenges, and that’s easier to integrate into your platforms. It also provides a competitive advantage that’s absent when you and your rivals are all beholden to the same service provider. To achieve unique results in the era of technology, you can’t keep strategising and thinking about data the same way. Instead, you need to consider the unique ways you can harness it to provide equally unique and bespoke solutions to consumers, who have now come to expect specific and customised offerings.

Dr Mark Nasila is chief analytics officer in FNB’s chief risk office