Bank loans: what if your call history was worth more than a payslip?
Over one in six people worlwide are excluded from the financial system. As they are unable to provide standard supporting documents, they can‘t obtain loans. In Africa, for example, artificial intelligence (AI) is boosting the creativity of financial institutions, which are developing new and sometimes surprising ways to prove their reliability. But at what ethical cost?
The way you charge your phone or how regularly you answer calls could one day determine your chances of obtaining a loan. For millions of people around the world, this idea is already a reality. It has become a flourishing practice in developing countries, where banking exclusion is extremely widespread. As many find it impossible to provide standard supporting documents like payslips or bank statements, financial institutions are breaking new ground by using “alternative” data to assess individuals’ credit standing. Digital habits, mobile phone behaviour, social media interactions and personality traits can all become indicators for creating a “financial signature”.
1,4 billion of adults without a bank account
Behind this transformation lies a fundamental issue: access to credit for those excluded from the financial system. In 2021, according to the World Bank, 1.4 billion adults worldwide didn’t have a bank account . Most live in rural areas, work in the informal economy, are young, are women, or live on the fringes of society. All these profiles are considered “risky” by conventional financial standards.
And there is an even more speaking figure: over 2 billion people work without contracts, payslips or social security cover (ILO, 2023). In these circumstances, how can an individual’s ability to repay a loan be assessed without going through the usual channels?
What Africa is borrowing from AI
The answer, as you can guess, lies with artificial intelligence (AI), with its ability to analyse unconventional data. In Africa, for instance, the rapid growth of mobile payments and digital connectivity is opening a whole new field of behavioural analysis.
- In South Africa, the agency TransUnion has launched Telco Data Score: a scoring tool based on phone usage habits, including call activity, number stability and recharge frequency. The result is a 25% to 35% improvement in predictive performance compared to conventional methods.
- Again in South Africa, African Bank is trying out the tool Worthy Credit , a psychometric test that assesses financial discipline, behaviour and personality traits. The questionnaire items ask participants to choose between two equally desirable statements. For example: Which phrase best describes you? “I manage my finances carefully” or ”I avoid risky financial situations” This method has made it possible to reintegrate into the system people rejected by traditional credit scoring, with reimbursement rates deemed satisfactory.
- In Kenya, the fintech company Tala takes analysis one step further: it cross-references up to 10,000 mobile phone data points, like internet browsing, geolocation and payment frequency, as a basis for granting microloans. Over 6 million loans have been provided through this method, with a default rate comparable to banking standards.
Ready for ethics?
However, this revolution raises some serious ethical issues. In 2019, a survey by ProPublica, a non-profit investigative journalism organisation, revealed that an algorithm used by an American lending platform systematically rejected applications from certain neighbourhoods, despite comparable financial profiles. This algorithmic discrimination, though unintentional, shows how biases present in historical data can be reproduced and heightened by machines.
The widespread use of personal data also raises questions about privacy, consent and transparency. In addition, the challenges from a technical standpoint are considerable. The quality, accuracy and representativeness of data are essential to avoid inaccurate assessments that could unfairly penalise borrowers who are in fact perfectly creditworthy.
Read also : Who is afraid of investing in Africa?
How far are we prepared to let algorithms decide an individual’s future? In a world where finance is becoming ever more digital, there is a strong temptation to replace human judgement with that of machines. However, true inclusion can only be achieved if we respect the dignity, complexity and potential of each individual .
In a world where data is the new currency, inclusive credit scoring could verywell be the key that unlocks the future for millions of people. But this power is a double-edged sword: it can open doors, but it can also create new ones.
