Share the article
twitter-iconcopy-link-iconprint-icon
share-icon
AfricaFebruary 1 2019

AI in Africa: a different kind of intelligence

AI has still to make its presence felt in Africa's banking sector, but brings with it opportunities to bank the unbanked and engage with the continent's youthful population. However, as Kit Gillet reports, Africa's unique needs mean achieving these goals is not straightforward.
Share the article
twitter-iconcopy-link-iconprint-icon
share-icon
Standard Bank

Banks around the world have long touted artificial intelligence (AI) as a potential game-changer that will revolutionise financial services, helping to reduce overall staff needs while offering clients greater options and improved and faster services.

While this potential is still far from being realised, many banks are already investing heavily in the technology and the required human resources. In Africa, AI and related technologies are also being seen as a way of opening up new markets and helping banks reach into the continent’s vast unbanked population.

“Africa still has the most severely underserved population in the world with respect to financial services, and it is the fastest growing demographic in the world,” says Andrew Watkins-Ball, CEO and founder of Jumo, which builds and operates technology platforms delivering digital financial services to individuals and small businesses, and which has offices across Africa. “The opportunity presented by AI in that context is to move past bias and create fairer and more responsive financial tools for emerging consumers.”

African potential

In Africa, banks and other financial institutions are increasingly looking at AI to provide ways to drive innovation, though with an immediate focus on the automation of repetitive back-office procedures, as well as in areas such as fraud detection and prevention. They are also looking to use chatbots to respond to predictable customer queries.

“Machine learning represents the most substantive opportunity in AI,” says Alpheus Mangale, chief information officer of South Africa’s Standard Bank Group, which over the past year has successfully piloted projects automating processes including the reporting of foreign exchange transactions to the central bank and large parts of the process of extending further credit to existing business clients. “We’ve seen AI’s capacity to scan vast amounts of data to identify patterns, draw conclusions and make predictions quickly and efficiently in relation to routine, data-based, predictable tasks.

“What can be achieved by combining the number crunching abilities of AI with human creativity, innovation, experience and empathy... that’s where we’ll see game-changing impacts.”

In a February 2018 report, McKinsey & Company declared that Africa’s banking markets are among the most exciting in the world, pointing to the continent’s overall bank-ing market being the second fastest grow-ing and second most profitable of
any region, while also being a hotbed of innovation.

However, the report also found that many African retail banks have excessive back offices, with staff often labouring under paper-heavy, manual processes, and with a higher proportion of staff in support roles such as IT, risk, marketing, human resources and finance. “Technologies such as robotics and AI can deliver striking benefits,” the report predicted. “Robotics can be used to automate most repetitive operational tasks at little cost, while AI-enabled chatbots can handle up to 50% of client communication in call centres.”

If banks were to automate the top 20 processes out of about 600, the report found, they would likely capture 45% of the benefits of digitisation.

A head start

Some involved in the African banking sector believe that AI is already having an impact on the continent’s financial services.

“AI has started permeating into intelligent organisations and advancing to the fundamental toolset for daily engagement with people for both clients and employees,” says Fred Swanepoel, chief information officer at South Africa’s Nedbank, which currently has more than 100 bots in production, with automated processes implemented across areas such as corporate and investment banking, wealth management and shared services.

“The application of AI technology by banks has started to gain momentum, raising the optimism of investors about the economic benefits of improved client engagement, cost savings and revenue enhancement for the incumbent players,” he says.

One key aspect of the African demographic that favours the use of AI is its relative youth, with younger populations more comfortable with and willing to embrace automation and digital-first financial services.

“Africa is building the youngest demography in the world and this young population will not be served with traditional banking channels,” says Austine Abolusoro, group head of online banking at Nigeria-based United Bank for Africa Group (UBA).

“This population want to chat instead of calling or sending e-mails. So we gave them a channel they are already used to,” he adds, pointing to the bank’s launch of a chatbot, Leo, in early 2017, which provides banking services to customers through social media platforms such as Facebook. “They can perform most of their banking transactions over the chat channel. This is a huge opp-ortunity for us, beyond what we initially imagined.”

According to Mr Abolusoro, more than 72% of subscribers on the platform are under 35 years old, with more than 1 million subscribers across 16 countries using three languages (English, French and Portuguese). Millions of transactions have so far been processed, more than 50 million conversations generated and thousands of staff hours saved. “Over 20% of calls through our call centre now go through the channel,” says Mr Abolusoro, adding that the banks also sees potential in deploying AI in credit scoring and consumer lending.

Reaching the unbanked

One major opportunity for financial service providers in Africa is the potential for AI and automation to help them expand into underserved markets.

According to those involved in the sector, banks and financial organisations looking to grow in Africa, or extend their footprint into new emerging economies in the region, are now able to use automation as a means of scaling up their operations in both a fast and agile way. This can be particularly helpful when it comes to avoiding the higher costs and longer timescale of setting up large-scale physical infrastructure, as well as bypassing issues related to finding a strong local team on the ground.

“I think fundamentally one of the problems we are solving is that any business, and I think this is particularly true in Africa, is constrained by its access to skilled resources,” says Terry Walby,
CEO and founder of London-based Thoughtonomy, which in October 2018 entered into a partnership with DigiBlu, a South Africa-headquartered company focused on digital workforce implementation, to deliver what they claim will be the first software as a service-based intelligent automation solution in Africa.

Mr Walby adds: “If you’re in one of the more remote African geographies is it possible to build a scale operation to run your financial processes, or do you end up having to outsource those to one of the bigger African countries such as South Africa, or even to an offshore location? What we are finding is by making processing more efficient you can do it with fewer resources, and therefore you can retain the delivery of business operations into some of the more [peripheral] countries that you wouldn’t otherwise be able to.”

There are other opportunities, notably the ability to use AI to reach some of those who have so far remained outside the financial system; in many African countries a significant percentage of the population remains unbanked. “Having this huge unbanked base is an opportunity to deploy technology that could help scale up the customer acquisition for banks,” says UBA’s Mr Abolusoro.

Sub-Saharan Africa is home to all 10 economies where more adults have a mobile money account than have a financial institution account: Burkina Faso, Chad, Côte d’Ivoire, Gabon, Kenya, Mali, Senegal, Tanzania, Uganda and Zimbabwe, according to the latest Global Findex, released by the World Bank, which measures financial inclusion and fintech adoption. Meanwhile, globally 20% of adults without an account at a financial institution reported lacking the documentation needed to open one, with the highest percentage in countries such as Zimbabwe (49%) and Zambia (35%), as well as the Philippines.

“AI platforms plugged into different data sources could help automate the identification process and bring a lot more of these people into the financial services domain,” says Mr Abolusoro.

Low-hanging fruit

Despite the promise of AI when it comes to opening up new financial services and providing fresh products to customers, the approaches to date across Africa have been similar to those taking place in other regions of the world, with cost cutting at the forefront of many efforts.

“There is a lot of talk about using AI to improve service quality and deliver more value to the customer, but at the end of the day banks mostly use it to cut costs, which is fine, but the benefits go to a very different group of people,” says Co-Pierre Georg, an associate professor at the African Institute of Financial Markets and Risk Management at the University of Cape Town.

There are also challenges related to the development of AI technology, and the inability to rely on importing solutions that have worked elsewhere in the world. “You can’t just take a European or US solution and drop it into Africa and see how it works. We have very specific challenges, for example remittances, which banks have a hard time solving,” says Mr Georg. 

Still, many see the increasing adoption of emerging technology such as chatbots and the automation of back-office processes as a gateway to utilising AI and automation in far more comprehensive ways across financial services.

“The first wave [of adoption] is about cost efficiency, the second is about improving business performance, and the third is about how you transform what you do, with a pool of digital workers that can do things you couldn’t previously,” says Thoughtonomy’s Mr Walby. “Pretty much everybody starts with wave one. It’s a problem begging to be solved, but also getting a quick return on investment really helps to accelerate the use of the technology.” 

Growing pains

One major issue that will govern how quickly financial service providers in Africa are able to scale up their automation and AI capabilities is the availability of human resources.

“There is a much smaller resource base in Africa of people who not only have training in an academic sense but also have tried-and-tested experience of algorithmic application at scale,” says Jumo’s Mr Watkins-Ball. “There are just far fewer people working on the problems, and the resources that do exist are scarce and highly fought over. This is a development challenge for Africa, which means that this kind of expertise, and the solutions, need to be imported. There is some local talent, but it is a very small fraction of the talent that is required.” 

Even so, there are some countries that are proving to be far more innovative when it comes to digital and AI innovation. “There are definitely some countries that are leading the way in Africa,” says Mr Watkins-Ball. “Kenya, as a result of M-Pesa [a mobile phone-based money transfer, financing and microfinancing service launched in 2007] is one, but there are also some examples that are extremely promising in places such as Ghana and Uganda. They have thriving digital payment environments that are growing rapidly and have more and more innovations and customer value attached to them. It’s an exciting time in that regard.”

Role of fintechs

The rise of fintech companies also presents a challenge for traditional banks and financial service providers, and there is a sense that without innovating (which includes the adoption of AI technology) banks will struggle to compete.

“The fintechs and big technology companies are eating into profit streams not only in payments, but increasingly also in the traditional credit parts of the financial sector,” says Standard Bank’s Mr Mangale. “Big tech companies such as Amazon, Alibaba and Tencent have networks, lots of capital and strong customer relationships. They are in a position to be able to stand between a bank and its customers,” he adds, saying that as a bank it is looking for ways to partner with fintech companies to grow opportunities and markets.

According to University of Cape Town’s Mr Georg, the biggest challenge for traditional banks is getting ready for this shift, where tech companies become hybrids. “This is going to be a megatrend, and African banks are particularly badly prepared as our universities don’t produce enough graduates that have the right skill sets. You can augment specific skills in house, but you can’t teach these broad foundational skills.” 

He also predicts that when it comes to fintech companies, what is likely to happen in Africa is banks will increasingly buy up small startups as soon as they become viable and attractive. “Instead of innovating in house they will wait to see what works in the marketplace and buy them up.” 

Still, there is a sense that African banks are aware of the challenges and are willing to embrace the opportunities that AI and automation offer.

In October 2018, First Bank of Nigeria (FBN) held a three-day Fintech Summit with the theme ‘The future of banking – the role of AI and big data’. “Technologies such as AI and big data need to become part and parcel of our business models across the board,” FBN’s managing director, Adesola Adeduntan, said at the event.

FBN is far from alone in believing this.

“In terms of awareness of what exists, what can be done, Africa is lagging some of the other geographies such as the US and parts of Europe,” says Mr Walby. 

“But what we’ve observed in terms of adoption is that there is a real openness to embrace technology in Africa. There is a real sense in Africa that if we want to compete on a world stage we need to be doing things differently.” 

Was this article helpful?

Thank you for your feedback!

Read more about:  Africa