While AI cannot completely prevent fraud in SA’s burgeoning iGaming industry, the technology can help stakeholders monitor and detect threats in real-time, making a significant difference, says Mariia Halaida, head of business development at iGaming software solutions firm Softswiss.
Softswiss has described SA as Africa’s most advanced and regulated gambling market, noting year-on-year (YOY) growth in the country’s iGaming industry.
In November 2025, Softswiss claimed that in FY2023/24, the country’s total gambling turnover reached €56.8 billion (about R1.13 trillion), representing a 40.2% YOY increase, while gross gambling revenue (GGR) stood at €3.05 billion (about R60.99 billion), up 25.7% from the previous year.
The betting sector dominates the market, accounting for 60% of total GGR, with online betting alone contributing €1.5 billion (about R29.9 billion) in revenue.
Miranda Guliashvili, head of regional growth at Softswiss, said: “Approximately 49% of total GGR now comes from online betting, with 81% of all betting revenue generated through mobile channels. This shift reflects the country’s high smartphone penetration rate (around 79%), affordable mobile internet and rapid uptake of 5G technology. Younger, mobile-native players (median age: 28.7 years) now represent the core audience, especially in urban centres like Johannesburg, Cape Town and Durban.”
Halaida says within the 2024/25 financial year, SA’s iGaming industry was valued at approximately R75 billion.
Increasing fraud
However, increasing fraud threatens to hamper this growth.
According to data from global identity verification and compliance technology firm Sumsub, SA recorded a more than threefold increase in iGaming fraud during the second half of 2025.
“The biggest problem with fraud, not just in South Africa, but worldwide, is that it is detected after it happens and not in real-time,” says Halaida, who adds that this is where AI can prove useful by analysing large datasets, spotting fraud and money-laundering activity, and immediately alerting organisations to such activity as it happens.
Halaida says organisations continue to view AI as a silver bullet, believing that simply investing in and applying AI technology solutions will address all challenges.
“AI is completely overhyped… but it cannot solve everything at once. You need to first identify what the problems are that must be addressed and then, using results that can be measured, apply AI,” she continues.
Regulation maturity
Halaida says growth potential, market value and size aside, Softswiss’ business case for regional expansion also depends on a region’s regulatory environment.
The Softswiss executive says that, on reflection over the past five to 10 years, SA’s regulatory environment has evolved.
“Regulations are becoming tighter but also clearer and understandable. We go where the path is clear because it is better to understand [the environment] from the beginning, rather than to try to spread to other countries and try to work there.
“We prioritise markets where we see that regulations are in place, and we understand those regulations, as well as how we can apply and adapt our software to comply,” she adds.
Sunday World online reports that SA’s Department of Trade, Industry and Competition is reviewing the National Gambling Amendment Bill, while the Advertising Regulatory Board is rechecking its Gambling Code to tighten control over marketing to minors.
However, stakeholders within the iGaming industry must wait to see whether these measures will add value.
Agentic AI: A double-edged sword
Mark Walker, director and co-founder of Transformation for Innovation (T4i), suggests agentic AI fraud detection and mitigation shifts security from passive, rules-based detection to autonomous, real-time investigation and action.
"While generative AI has allowed fraudsters to scale deepfakes, synthetic identities and polymorphic phishing attacks cheaply, agentic AI counteracts this by deploying specialised chained agents that focus on complex workloads across multiple checkpoints and sources," he explains.
"These agents analyse complex data silos, adapt to new fraud typologies without prior data labelling, and execute multi-step containment workflows within milliseconds."
However, Walker cautions that the technology is not without risk.
"While this is highly effective at stopping advanced, machine-speed fraud, deploying autonomous chained agents introduces unique operational, technical and financial risks. Shifting final decision-making power from static rules to autonomous software creates a new set of vulnerabilities, including untraceable decision paths, high token costs and system-wide automation errors."
Looking ahead
Going forward, Softswiss is adamant that its market experience, platform technology and scalability will add value to the African and South African markets.
Halaida suggests that, in much the same way smartphones were introduced as “the next big shiny thing” and eventually became an established resource within society, AI will grow to emerge as a common and widely used technology resource.
Share
