It seems as though South Africa has tried everything to tackle rampant tender corruption. However, the fight against corruption resembles a tragic opera with a familiar plot: a scandal breaks, public outrage follows, the president establishes a commission of inquiry and then … not much happens after that. Then the cycle is repeated.
We have a forest of oversight bodies all armed with various laws and enforcement powers. Yet dodgy tenderpreneurs continue to beat the system, while the infrastructure they are contracted to fix crumbles. It’s time to admit a hard truth: South Africa’s human-only oversight system has failed to tackle tender corruption.
It’s not that the good people in these oversight bodies aren’t trying; it’s just that they are going into a gunfight armed only with a penknife. The corruption networks are too complex, the data too vast, the connections too well hidden.
Clearly, we need a new kind of public servant. One that doesn’t get tired, doesn’t get bribed and has no “cousins” needing jobs.
It’s time to appoint Minister Tender AI (Artificial Intelligence). Do I hear any objections?
Some might say: “An algorithm can’t understand nuance”, or dismiss it as a “dehumanising technocratic fantasy”. But these arguments miss the point entirely. Using AI is not about replacing human judgment; it’s about arming it with superhuman insight. Think of it not as a cold robot overlord but as the most powerful, unblinking watchdog imaginable.
Its portfolio would be simple: to serve as the central nervous system for every cent of public procurement, creating a public dashboard that shines a light into the darkest corners of our tender system.
The blueprint already exists, and not in some Scandinavian utopia. Look at Albania; tired of the same graft that plagues South Africa, enterprising Albanians launched Diella AI, a “forensic digital detective”. Diella consumes billions of data points: tender documents, company registries, financial records, news feeds; and maps the hidden networks of influence that human auditors would take years to uncover.
Diella can flag a tender with suspiciously narrow specifications tailored for one company. It can instantly reveal if the winning bidder’s director is related to the municipal official who approved the contract.
For the first time, Albanian citizens can have some assurance that their taxes will actually be used to construct a road rather than a private mansion.
This is the key solution we are missing. A South African Minister Tender AI would work 24/7 like Diella. It would integrate data from Sars, the Companies and Intellectual Property Commission and banking records. A company that consistently wins tenders but reports minimal profits to Sars? Flagged.
A “new” bidder that shares a bank account with a banned contractor? Exposed. Those complex webs of front companies designed to obscure ownership would be rendered useless in seconds.
Those complex webs of front companies designed to obscure ownership would be rendered useless in seconds
By analysing bid prices across the country, the AI could spot the telltale signs of bid-rigging, such as identical, inflated quotes or a rotating circle of winners, patterns that are invisible to the naked eye but glaringly obvious to a machine.
The most powerful feature, however, would be the public dashboard. Imagine a single, open website where any journalist, activist or ordinary citizen could track every government tender in real time.
You could see who bid, who won, at what price, and crucially, the AI’s calculated risk score for that tender. Suspicious activity would be immediately flagged.
This doesn’t just aid investigators, it democratises oversight, turning every citizen into a guardian of public funds.
Of course, Minister Tender AI is not a magic wand. Challenges will exist because AI is only as good as the data it’s fed, and government databases are not always up to scratch. There’s a risk of algorithmic bias if it learns from our corrupted past. And corrupt actors will adapt, finding new ways to hide their tracks.
This is why the “human-in-the-loop” is non-negotiable. AI’s role is to flag the anomaly; the prosecutor’s role is to build the case.
In that regard, AI provides the evidence and the court delivers the judgment. The real obstacle, as always, is not technical; it’s political.
A system this transparent would be a direct threat to the powerful networks that profit from the status quo. Expect tender mafias and their enablers in government to fight it tooth and nail.
Implementing Minister Tender AI would be the ultimate test of political will. The question is whether we have the courage to use it.
• Lourie is the founder and editor of TechFinancials
.
