South Africa accounts for the largest share of deepfake-related identity fraud in Southern Africa, according to a new report highlighting the growing role of artificial intelligence in digital scams.
The findings were published in the 2026 Digital Identity Fraud Report by Smile ID, which examined biometric verification trends and identity fraud patterns across the region.
Deepfake technology driving new fraud tactics
The report found that deepfake-driven fraud represents about 22% of identity fraud incidents in Southern Africa originating from South Africa, the highest share among countries in the region.
Deepfakes involve the use of manipulated images, video or audio created with artificial intelligence to impersonate another individual. In digital identity verification systems, fraudsters may attempt to pass biometric checks by presenting altered or AI-generated media.
Smile ID said that biometric-related fraud dominates in South Africa’s digital verification systems. According to the report, nearly nine out of ten verification attempts rejected due to suspected fraud involve impersonation or spoofing during biometric authentication.
The largest category identified in the research was “no-face-match” impersonation, accounting for about 47% of detected fraud attempts. In these cases, the person attempting verification cannot be reliably linked to the identity they claim.
Such cases are often associated with stolen identities, the use of intermediaries or “mule” accounts, or abuse occurring after legitimate account registration.
Another major category is spoofing, which accounts for roughly 40% of fraud incidents. Spoofing refers to attempts to bypass biometric security systems such as liveness detection and facial similarity checks.
Fraudsters may use techniques including face-swap technology, manipulated photos, or deepfake-generated videos to trick verification systems.
Document-based fraud, such as forged or manipulated identity documents, represented about 13% of cases. While still present, the report described these methods as secondary compared with biometric impersonation.
High-profile deepfake scams raise concern
Authorities in South Africa have also warned about deepfake scams involving well-known public figures.
The Financial Sector Conduct Authority (FSCA) previously issued a warning about manipulated videos and images impersonating prominent individuals, including President Cyril Ramaphosa and businessman Patrice Motsepe.
The fake material promoted fraudulent investment schemes and online trading platforms falsely claiming association with the two figures.
One widely circulated deepfake video appeared in March 2023 on the social media platform X. The video showed what appeared to be President Ramaphosa announcing plans to demolish the Voortrekker Monument and Pretoria’s Loftus Versfeld rugby stadium.
Observers later pointed out signs that the footage was manipulated, including minimal body movement and unnatural facial expressions.
Experts believe the clip was likely created using early artificial intelligence voice-cloning and video-editing tools that were widely available online at the time.
Beyond political impersonation, scammers frequently mimic major companies to gain victims’ trust. Fraudsters may pose as representatives of banks, telecommunications companies, or well-known brands such as DStv in emails, text messages or phone calls.
These messages typically attempt to convince victims to pay fraudulent fees or disclose sensitive personal and financial information.
As AI tools continue to improve, cybersecurity specialists warn that deepfake scams may become increasingly sophisticated, making public awareness and stronger digital verification systems essential.
