Nearly 60% of South African organisations have noticed a surge in financial crimes facilitated by AI technology, higher than the global average (56%). This has resulted in a sharp increase in the cost of compliance for financial institutions.
This is according to the LexisNexis Risk Solutions true cost of financial crime compliance study, which found that the price tag of compliance at South African financial institutions is estimated to be at $1.4 billion. This includes labour, technology, outsourcing and infrastructure costs.
Daniel Bisson, director at LexisNexis Risk Solutions, stresses that financial institutions must prioritise financial crime compliance to avoid costly fines and reputational damage. By adopting advanced compliance solutions, organisations reduce false positives, streamline due diligence, and improve customer experience, he says.
Key trends in 2025
The study identified key trends that will impact financial institutions.
- 2025 will see AI begin to bear fruit in the fight against financial crime, says Bisson. “With a 56% rise in AI-driven financial crimes, banks are reassessing existing tactics. A combined approach using AI tools, human expertise, and quality data is seen as the best solution to counter these threats."
- Globally, financial institutions spent an estimated $34.7 billion on financial crime compliance technology in 2024.
- The growing importance of private-public partnerships (PPPs).
“The rise of complex criminal networks and transnational financial crime has driven a shift toward strategic collaboration among regulators, law enforcement, and the private sector," says Bisson. "By sharing intelligence, these partnerships strengthen efforts to detect and prevent money laundering, terrorist financing, and related financial crimes.”
He refers to South Africa’s 'project blood orange' as an example. The project saw the South African Police collaborate with KPMG and the South African Revenue Service to investigate a criminal network responsible for the country’s rhino poaching crises. As a result, 16 people were arrested for money laundering, corruption and conspiracy, and rhino poaching decreased by 30%.
Customer expectations
LexisNexis Risk Solutions study also notes that customers will continue to set the bar higher.
“The digital transformation spurred by Covid has continued, raising consumer expectations for fast, secure, and personalised services with minimal hassle. Customer experience is now the number one driver of customer loyalty, making it crucial for financial service providers to balance seamless service with fraud prevention and compliance –a challenge that will remain a key focus.”
Bisson adds that the expansion of telecommunication and network service providers into financial services offers consumers numerous benefits, including more choice and improved access to financial services.
“However, this growth is not without risk. Bad actors are always on the lookout for opportunities to exploit, particularly when new players enter the market. New providers must ensure they have the same level of financial crime compliance controls that traditional financial institutions possess.”
Growing risks
Mark Walker, VP of data & analytics at IDC, says the reality is that AI enhances fast, complex, and innovative computing across many domains, but as it integrates into business processes, AI-driven financial crime will also rise.
"Since South Africa is already a significant hub for African and global financial transactions, we can expect the country to be targeted for AI driven financial crime.”
Financial technology suppliers are aware of these increased risks and are hardening their solutions to reduce the opportunity for these events and also to comply with stringent GRC requirements, leading to higher R&D costs passed to customers.
“The Chinese DeepSeek model is already challenging OpenAI by providing alternative AI solutions at 1/1000th of the cost – this trend will continue and accelerate rapidly as more competitors come to market and the cost per token falls dramatically,” he says.
“While this example is at the macro level it will flow down to individual AI applications so GRC costs should ideally decrease over time as manual processes are automated and operational quality and speed improves. However, AI is not infallible so human-based checks and balances will still be required to 'watch the watcher' and ensure that issues around misinformation, manipulation, transparency, local regulatory nuances do not occur.”
Share