Africa’s fast-growing mobile money industry, now worth about $1.4 trillion, is facing a rising wave of fraud powered by artificial intelligence. Experts warn that the same systems helping millions gain access to financial services are now being targeted by more advanced and organised cybercrime.
Data from GSMA shows that Sub-Saharan Africa handles nearly two-thirds of the world’s mobile money transactions. The region has around 1.2 billion registered accounts, with 341 million people actively using these services.
But as usage increases, so do risks. In an interview with IT Web Africa, Thalia Pillay said the industry is entering a new phase where fraud is becoming more widespread and harder to detect.
She described mobile money as a major success story for financial inclusion, but warned it is also becoming a key target for criminals. Industry figures show that 90 percent of mobile money providers reported identity fraud in the past year, while 88 percent faced social engineering attacks. Across the continent, cybercrime is estimated to cost more than $4 billion each year.
In Zimbabwe, the impact is already clear. The country’s ICT minister, Tatenda Mavetera, said mobile money fraud costs the nation over $30 million annually. She added that phishing attacks have risen by more than 40 percent, targeting the country’s expanding digital economy.
Mavetera warned that as more services move online, new risks continue to emerge. Each step towards digital growth, she said, opens another door that criminals can try to exploit using AI tools.
According to Pillay, fraud methods are changing quickly. Criminals are now using techniques such as SIM swap attacks, fake identities created with AI, and deepfake scams where voices are copied to trick business owners into approving payments.
She explained that AI can now produce fake documents and even biometric data that can pass identity checks, known as Know Your Customer processes, at scale. This makes it much harder for financial institutions to spot fraud early.
Another problem is that many fraud detection systems used in Africa are built for other regions. These tools often assume stable internet access, strong identity systems, and consistent user behaviour. In many African markets, where people may share devices or use multiple SIM cards, these assumptions do not always hold.
Pillay said this mismatch means systems can either block too many genuine users or fail to stop fraud effectively. She stressed the need for solutions designed specifically for African conditions, using local data and patterns.
The growth of cross-border and interoperable payments is also adding to the challenge. As money moves more easily between banks and mobile wallets across countries, criminals are finding new ways to move stolen funds quickly, often faster than regulators can act.
Pillay noted that fraud networks operate across borders, while many defence systems remain limited to single countries. This gap, she said, gives cybercriminals an advantage in an increasingly connected financial system.
