Uganda among top targets of AI-generated fraud 

Artificial intelligence is the tool of choice for impostors attempting and executing fraud, with East African countries not spared, as Tanzania ranks top in the region while Uganda and Kenya are not far behind, among countries that are hit by groups that invest as little as $1,000 but inflict losses of up to $2.5m a month on businesses, according to the Identity Fraud Report 2024, released Tuesday. 

The report, compiled by London-based digital verification platform Sumsub, puts Tanzania at 5.39 percent average fraud rate - well among the top five countries on the continent - with Uganda and Kenya coming in at 4.53 percent and 4.01 percent, respectively. 

Of concern, the researchers say this year’s rate for Kenya, is more than double its fraud incidence over the past three years, meaning the country is fast spiraling in the wrong direction, as AI and other technology advances make it increasingly cheaper to commit fraud. 

But Nigeria trumps other countries on the continent, with the highest average fraud rate of 5.91 percent this year among all verifications – this having doubled from last year – followed by Algeria (5.78 percent), Tanzania, Madagascar (4.92 percent) and Chad (4.76 percent) taking up the top five slots of fraud rates. 

For financial gain, fraudsters target individuals and businesses under what is called identity fraud and account takeover where a hacker breaches systems and gains access to the victim’s online account and data, without their consent. 

“In today’s digital world, identity fraud poses a serious threat to individuals and companies. In 2024, 67 percent of firms reported a fraud increase,” says Andrew Sever, co-founder and chief executive officer of Sumsub. 

According to the report, account takeover attacks have become one of the most damaging forms of fraud, which from 2023 has surged by 250 percent year-on-year, a staggering growth compared to a 155 percent increase in the 2022/23 previous review period. 

The top five sectors and business categories most affected by identity fraud in 2024 are dating services and platforms with a fraud rate of 8.9 percent, online media at 7.7 percent, banking and insurance scoring 2.7 percent, video gaming at 2.3 percent and crypto at 2.2 percent. 

The report paints a worrying picture for African businesses, warning that in 2025 companies will continue to bleed money as fraudsters will increasingly rely on a broad range of AI-generated deceptive tools that are harder to detect. 

The grim picture draws on the experience that businesses have had, losing approximately $300,000 per fraud incidence in 2024, with nearly half of the companies and end users surveyed worldwide reporting being victims of identity fraud at least once. 

“Fraud-as-a-service is becoming a growing threat to African businesses, with identity fraud across the continent increasing at an alarming rate. The average year-on-year fraud rate growth in Africa has reached 167 percent,” the report reads in part. 

Average fraud rate across Africa 

However, Sumsub researchers noted that cases of deepfake fraud - where attackers deploy manipulated images, videos, or voices to impersonate individuals – have spiked across Africa by an average growth rate of 393 percent year-on-year, warning that AI-powered fraudster economy will explode next year. 

“In 2025, fraudsters will increasingly rely on AI not just for deepfakes - which are, basically, just a tip of the iceberg - but for a broader range of deceptive tools, such as AI-generated identity documents, real-life videos, synthetic voices, and AI-driven chatbots that impersonate real users. These innovations will make fraud harder to detect and more versatile,” said Pavel Goldman-Kalaydin, head of AI and machine learning at Sumsub. 

He further noted that to fight AI-powered fraud, businesses need to deploy advanced multi-layered solutions that can analyse and detect fraud across multiple vectors, ensuring that identity fraud is tackled not just at the visual or biometric level, but across all touchpoints.

This year, there has been a significant increase by four times in the number of deepfakes detected globally across all industries year-on-year.

Early this year, former Bank of Uganda executive director for supervision Tumubweinee Twinemanzi warned that whereas it would be difficult to protect Ugandans against AI-generated fraud, it was important that businesses arm themselves with “information on how to protect themselves on the technology they use,” noting that the Central Bank was ensuring that financial institutions submit look at various risk categories as part of operations.

Democratising fraud

DuApart from fraud, the Sumsub report also indicates that AI and deepfakes have been widely deployed to spread misinformation, citing the example of AI-generated images of “Disney World underwater” in which social media some people were hoodwinked social media users when they create an impression that Disney World in Florida was flooded after Hurricane Milton. 

Digital experts say the three primary factors reshaping digital crime are the advent of fraud-as-a-service platforms and easily accessible fraud tools, which they dub “democratisation of fraud”, but also digitisation of services as businesses transition online, as well as low levels of digital hygiene where users forego basic cybersecurity practices such as strong passwords and multi-factor authentication access. 

[email protected]