/ 10 April 2024

AI ups the ‘anti’ on money laundering

Bugs 4

Money laundering is once again a harmful headline in South Africa, with the speaker of the National Assembly charged with corruption and money laundering worth over R4 million. It’s a thorn in the side of the economy. Our country lags at 83 out of 180 on the Corruption Perceptions Index by Transparency International, a ranking that reflects how financial crime continues to undo efforts from the public and private sectors to claw us from greylisting.

At its core, money laundering is the process of concealing the origins of illegally obtained funds to make them look legitimate. This not only distorts financial markets but breaks trust in institutions and undermines regulatory frameworks – and the issue is not only economic. While money laundering is the currency of crime, it is often the linchpin for other criminal activities, intensifying social inequalities and threatening national security.

The link between fraud and money laundering adds complexity to the issue. Fraudulent activities, whether perpetrated through traditional means or cyber-enabled schemes, generate illegal proceeds that are laundered to hide their origins. The proliferation of digital channels and crypto-currencies has further exacerbated this challenge, providing anonymity and facilitating cross-border transactions. As fraudsters adapt their tactics to exploit vulnerabilities in financial systems, the need for proactive measures becomes paramount.

To effectively tackle this pressing issue, neutral, ethical and innovative solutions are required. Putting generative AI and robotic process automation (RPA) to work in organisations, from governments to banks and other financial service providers, can contribute to disabling this underworld – potential dirty transaction by potential dirty transaction.

Activated AI to deactivate the risks

AI has a good deal more to offer when it comes to devaluing the currency that funds the system behind money laundering.

For a start, it offers unparalleled insights into illicit financial activities, serving automation benefits that are exponential. These range from empowering an entity or task force to identify suspicious activities in real-time – and on a large scale – to detecting transaction patterns and identifying anomalies. Machine learning algorithms, for instance, can sift through terabytes of transactional data in real-time, flagging suspicious behaviours and minimising false positives.

Natural language processing (NLP) algorithms analyse unstructured data sources, such as social media posts, for signs of potential money laundering activities. Similarly, network analysis techniques help find the links between transactions and entities or individuals behind such moves, bringing them into the light. Moreover, AI-powered solutions can adapt to evolving threats, learning from the past to enhance predictive capabilities.

The collective fight against money laundering is a matter of rigorous compliance and keeping financial institutions in check. Notably, ITWeb reports the lead is being taken by South Africa’s big four banks in the adoption of AI. For another regional bank with 13 subsidiaries on the continent that faced a penalty for non-compliance in its country of domicile, RPA proved invaluable to lift compliance through rapid data cleaning, monitoring, verification, alerting and tracking actions taken around money laundering risks.

HSBC is also said to use AI to screen over 1.2 billion transactions for signs of financial crime every month. The bank indicates that it now identifies “two to four times as much suspicious activity as the previous system” with more accurate risk detection.

Despite its potential, the adoption of AI in finance faces barriers.

Chief among these are concerns of data privacy, governance and ethics, and algorithmic bias. Financial institutions must navigate complex regulations governing the collection and processing of customer data, ensuring compliance with stringent data protection laws such as the General Data Protection Regulation (GDPR) and US Health Insurance Portability and Accountability Act (HIPAA).

Data privacy and security concerns loom large, as AI is in its operational infancy where rules and security standards have yet to be fully established. In a market such as South Africa, we need more investment to secure a skilled workforce that is capable of implementing AI solutions; a human force with formidable expertise in data science, machine learning and related fields.

Ethical concerns also arise as AI systems can perpetuate biases present in historical data, leading to unfair or discriminatory outcomes. Local barriers linked to this are seen in a lack of access to accurate local datasets. These barriers are compounded by a reliance on outdated technology and an entrenched resistance to shifting mindsets. As an industry, we further need to consider and learn from governance frameworks such as the EU AI Act.

Despite these challenges, AI can definitively up the ‘anti’ in the fight against money laundering if regulatory frameworks are adapted to accommodate technological advancements, and stakeholders collaborate to address systemic development. Right now, it is up to us to protect and enforce integrity in this development, to make sure we don’t up the ante on the systems behind our AI of the future.

What a time it is to make AI light work.