Listening to evidence led in the Constitutional Court about a cartel of banks accused of manipulating the rand, I realised the significant legal and financial implications of this case. Heard last week, it focused on alleged collusion among local and international banks to manipulate the rand through collusion in the currency market.
The South African Reserve Bank and the competition authorities also investigated the impact of alleged manipulation through increased costs for businesses, consumers and the overall economy.
When the proceedings ended for the day, my evenings were immersed in reading Geoff White’s riveting new book, Rinsed: From Cartels to Crypto: How the Tech Industry Washes Money for the World’s Deadliest Crooks. White takes readers on a global journey to expose a new cyber super cartel. Through harrowing first-hand accounts, he reveals how tech has become the ultimate enabler of financial crime, with devastating human costs.
Now, as artificial intelligence (AI) integrates into South Africa’s digital economy, we stand at the face of a new era of possible collusion, one that could make the banking cartel look like child’s play. Many people in South Africa think “AI” is still in the distant future. But this perception is dangerously misleading. AI is already here, embedded in the algorithms that determine your loan eligibility, your job applications and even the advertisements you see online.
AI is not some far-off futuristic concept; it is a present-day reality with profound implications for fairness, opportunity, and economic justice. In South Africa, where historical inequalities continue to shape socioeconomic outcomes, the risks are particularly serious. AI systems learn from available data, and if that data reflects past biases — as it often does — the algorithms will inevitably perpetuate and even amplify discrimination.
Imagine an AI-powered hiring tool that systematically overlooks certain candidates or a loan approval system that disadvantages people based on demographics. This is already happening in other parts of the world, and South Africa is ripe for similar outcomes.
What keeps me up at night, however, is not just biased algorithms; it’s the possible emergence of AI cartels. Much like the banks accused of manipulating the rand, corporations could use AI to collude in ways that are undetectable and unpreventable by current regulatory frameworks.
Consider algorithmic price-fixing: AI systems can monitor competitors’ prices in real-time and adjust their own accordingly, creating de facto cartels without a single email or meeting. Such digital collusions leave no paper trail, making them nearly impossible to prove in court.
In a country where economic concentration is already high, the rise of AI-driven monopolies could crush competition, inflate prices and exclude small businesses from the market entirely. Moreover, data monopolies are becoming a new form of economic power. Large corporations with access to vast data sets could use AI to entrench their dominance, shutting out innovators and entrenching inequality.
When a handful of companies control the data and the algorithms, they control the future
When a handful of companies control the data and the algorithms, they control the future. Some might argue that the solution lies in government intervention. But let’s be realistic: our policymakers are already playing catch-up. The pace of technological innovation far outstrips the ability of legislators to draft effective laws. Worse still, the government itself lacks the technical expertise to regulate AI meaningfully.
This is not to say that regulation has no role to play. But premature, ill-informed laws could stifle innovation without addressing the real risks. Instead, the responsibility must first fall on the industry itself. The private sector must take the lead in developing transparent, accountable, and equitable AI frameworks. This means committing to ethical practices such as auditing algorithms for bias, ensuring diverse representation in AI development teams and prioritising transparency in how decisions are made.
Create an independent panel of commissioners with an AI framework, code of conduct and ethics. Violators must face penalties on percentages of revenue or fines on profit, including naming and shaming in the media.
It must not be a toothless body.
The commissioners must be chaired by a judge or legal person and the board must consist of industry-related competency skills, including finance, innovators and tech expertise. The competition commission and appeals court will never catch AI thugs because these are algorithms: the crime is committed by machines. Companies should also invest in skills development and inclusion initiatives to ensure that AI benefits all South Africans, not just a privileged few.
This is about actively creating a future where tech serves humanity.
• Lourie is the founder and editor of TechFinancials
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.