Deepfakes on steep rise in South Africa, fintech & banking hardest hit

The Senior Director of Fraud Product Management at TransUnion Africa, Amritha Reddy, has bemoaned the supersonic surge in deepfake cases in South Africa, pointing out that fintech and banking are among sectors with the biggest hits.

Speaking on a podcast, Reddy said their findings showed deepfakes had risen by 1,200 percent in the last year, with fake voices, images, and videos appearing impossible to identify.

The executive said the deepfake phenomenon, which has triggered what she calls a trust crisis, is driven largely by increasing digitalization in the country. “Internet penetration is at about 79 percent, with more than 50 million users and 124 million mobile connections, which is a fertile ground for exploitation. Added to that are 27 million social media users, and it’s clear that access, availability, and modern connectivity are amplifying this threat,” Reddy said.

Explaining why fintech and banking are the most vulnerable sectors, the TransUnion Africa official pointed to SIM-swaps and digital-first platforms.

She also mentioned the insurance sector as one of the prime targets and that, she said, happens through “altered claim videos and fake medical records.”

“Retailers face fake storefront scams. The media and government have been hit particularly through voice impersonation and political disinformation. So, it’s multiple sectors being exploited, especially in environments where digital access is widespread,” she added.

TransUnion has been using its Synthetic Fraud Model as a tool to catch synthetic identities scanning through public data.

A global problem

Reddy said while the deepfake problem is rife in South Africa, it is a major global issue which, apart from using AI to fight it, also requires a better understanding of what it is, how it manifests and how to better build security layers.

“We must understand the threat, where it comes from, how it’s exploited, and how to defend against it. For businesses, that means deploying the right technology to detect what’s real, synthetic, or a blend of both. Real-time anomaly detection tools, biometric authentication, and liveness checks are critical for verifying genuine faces and voices,” she recommended.

“This must be supported with device fingerprinting, behavioral analytics, and consortium data sharing across banks, fintechs, telcos, government, and the private sector so that such threats can be identified and stopped early,” she added.

Speaking further, Reddy cited a survey they conducted which highlighted consumers’ desire for stronger consumer security even at the expense of price. “Our recent Consumer Pulse survey found that 88 percent of South Africans would prefer stronger security on digital platforms over better pricing or product quality. What really matters to them is how their personal information is handled, trusted, and protected.”

A TransUnion research in 2023 showed most US consumers feared identity fraud.

Cultural awareness, digital literacy

Beyond the use of AI tools to catch the AI-powered fraud in deepfakes, Reddy proposes other measures.

“Equally important is cultural awareness. Africa is a very trusting society, and that trust can be exploited. So, beyond technology, we need smart business strategies, awareness campaigns, and digital literacy,” she said.

She added: “Digital literacy is especially critical for our youth. While digitally savvy, they can still be naive and therefore vulnerable. Building skills and competencies to support AI-based verification tools is essential. It’s a combination of technology, strategy, and education.”

Role of government

According to the Reddy, the government also has an important role to play. “Regulation and compliance are crucial. We need clear laws against abuse, with strict enforcement and strong penalties. Regulators should also modernize compliance frameworks to embed AI-related risks,” she says.

“Currently, many regulations are outdated and built for traditional threats. But we’ve leapt into a fast-moving digital world, and the regulatory environment needs to catch up quickly. AI itself should also be embedded into the regtech universe.”

Collaboration, she mentioned, is also key, adding that “we also need cross-sector and cross-border collaboration [because] fraud doesn’t stop at national borders.”

“Consortium data, intelligence sharing, strong law enforcement, and continuous engagement are key. It’s not just about gaining insights, but about turning those insights into action. We need a holistic framework that addresses people, processes, systems, businesses, and society.”

Article Topics

AI fraud  |  biometric authentication  |  biometrics  |  deepfake detection  |  deepfakes  |  financial services  |  regulation  |  South Africa  |  TransUnion