Why banks can’t leave decision-making to AI

AI is changing the digital banking outlook in West Africa, creating a new breed of customer-focused and innovative products. There are AI solutions applied to meet some of the most pressing challenges for banking in the region: financial exclusion, fraud and operational inefficiency. Credit Scoring and Loans. And here is a key application of AI: credit scoring. Through ML algorithms, analysing alternative data sources including mobile phone usage and utility bill payments, banks and fintechs are able to make lending decisions on individuals and businesses who were not traditionally part of the formal banking system. This has created a new market for microloans and other financial products and has contributed to advancing financial inclusion in the region. Another important customer for AI is customer service. Chatbots and virtual assistants built on artificial intelligence (AI) that act as customer service representatives are being used to offer round-the-clock support to consumers, direct them to relevant pages of the banks’ websites or even automate simple banking services, such as checking of balances and transaction histories. This has led to a better customer service experience and has lowered the cost of providing customer support.

AI-based chatbots and virtual assistants are now growing popular in West Africa’s digital banking industry. Organisations are using these technologies to offer 24-hour customer service and automate repetitive tasks, such as balance checking, transaction history access, and dispute resolution. This has contributed to better customer experience, as well as decreasing the amount of time and people working in customer service. With the power of AI, chatbots are able to comprehend natural language and have fun evolving conversations with customers. They can even be combined with WhatsApp to deliver a customer service experience that’s both frictionless and more convenient. In Kenya, for example, a microfinance bank provides an AI-based chatbot on WhatsApp where customers can apply for loans, check account balances and seek financial advice on the move. While, in other words, small independent vendors or manufacturers lack internal resources and funds to invest in AI smart assistants helping them out with customer service (e.g., sales), big banks or fintechs with too large a data donor have an understanding of how much people can manage, leaving aside companies; then they would rather go for more complex problems.

AI is being deployed to help meet some of the major challenges facing the banking industry in West Africa: financial exclusion, fraud and operational inefficiencies. Using the force of AI, banks and fintechs are creating all sorts of disruptive technologies aimed at changing the regional financial landscape. One of the biggest obstacles is economic exclusion. Most of the population in West Africa remains unbanked and does not have access to even basic financial services. AI is working to solve this problem by allowing the creation of alternative credit scoring and lending models that can evaluate the credit eligibility of people and companies who were once excluded from the formal banking economy. Another key challenge is fraud. The emergence of digital banking has resulted in a surge in online fraud, with lasting consequences for customers and banks. This is where AI comes in, as it is empowering the creation of more advanced fraud detection and prevention mechanisms. The third major obstacle would be ineffectiveness in operations. Most West African banks are dependent on legacy systems and manual operations that can slow down transactions, as well as make them more expensive and prone to error. AI is working to solve this by automating even more back-office tasks, but the efficient speed with which businesses now operate means that we are left looking for ever more things to automate in order to save time and money.

If data is the blood of AI, any solution that runs on incomplete or wrong data will be no good. There are several obstacles to data collection, management and governance in the digital banking space across West Africa. One of the major hurdles is that the data infrastructure isn’t standardised. From legacy systems to lending: Many banks and fintechs rely on technology from 20 years ago that makes it hard for them to collect and analyse data from various sources. Privacy and security of user data are other problems to be addressed. The more we rely on AI, the more apprehensive we become that our data will be abused or fall into the wrong hands. To overcome these obstacles, banks and fintechs will have to ensure they have a strong data management infrastructure in place. It should include an explicit data strategy, shared company-wide data infrastructure and standards and policies for how to maintain privacy as well as secure user data.

AI has the capability to revolutionise digital banking in this part of Africa, but it’s not a magic wand. AI algorithms are only as good as the data they are trained on and can be biased or fallible. That is why we need a human-in-the-loop to validate that AI-based decisions are fair and correct, both from the ethical and legal point of view. This is especially crucial in high-stakes domains (e.g., credit scoring and lending), where an erroneous decision can have a life-changing consequence for the individual. There is also a stronger recognition in the region of the need for human oversight, regulators say. Nigeria’s Central Bank, Kenya’s Central Bank, and Ghana’s Central Bank have issued new directives on AI and digital risk, which come with requirements for qualitative checks around risk reviews, audit trails, and human intervention in high-impact decisions. Through this human oversight, the banks and fintechs can ensure that they are deploying AI in a responsible manner, which upholds ethical standards, securing trust with their customers in return.

User attitudes and acceptance of AI in banking are key determinants for the deployment of AI-based solutions in the West African banking sector. Even though AI can make a dramatic difference in the digital banking industry, people need to always keep in mind that this is a technology with which many users aren’t yet comfortable or familiar. And there is a need to establish trust and engage in an ongoing process of educating users about the potential benefits of AI. There’s certainly a digital literacy challenge that needs to be addressed. Limited customer digital education is the biggest barrier to activating digital services, according to 77.3% of 203 senior banking executives from 40 African countries in a poll. In response to this challenge, banks and fintech companies have to invest in financial literacy programmes and communicate clearly with users about how AI functions. It’s also critical to be transparent about how we are using AI and to provide users with the ability to control their data. Banks and fintechs can help to promote the adoption of AI-supported solutions by fostering trust and knowledge amongst users, with the potential to make the financial system fairer and more inclusive through these innovations.

Sola Longe-Okenimkpe is the COO of Nuvu Africa, specialising in digital transformation and financial ecosystems with over 30 years of experience in diverse sectors across Nigeria and Central/West Africa. She has expertise in FinTech, banking, telecommunications, and hospitality, holding various roles in marketing communications, brand management, sales, public relations, and e-commerce.

Join BusinessDay whatsapp Channel, to stay up to date