SITA, Trapped Between Responsibility and Authority.

The first problem is that the customer, in SITA's model, is the department. Not the pensioner whose grant did not arrive because CPS and SASSA were still arguing about who had lawful authority to run the payment system. Not the new parent whose child's birth registration is trapped inside a HANIS replacement that has been “nearly ready” for the better part of fifteen years. Not the accused in a criminal matter whose trial is postponed because the digital recording system went down and no one can agree whose contract was supposed to be maintaining it.

Customer-centricity, in a procurement frame, means making the departmental client happy in the moment of transaction. Citizen-centricity, which is the thing the Constitution actually asks for, means being accountable for the continuity of the service the citizen never sees and only notices when it breaks. SITA's copy cannot tell the difference between those two things, and that confusion is not decorative. It is the reason a payment can lapse, a court recording can stall, a forensic system can sit under interdict, and no one inside the organisation feels that the person at the end of the chain is, strictly speaking, their customer at all.

The judicial record sits at the exact opposite end of the word “innovation” from where SITA's About page places it. What the courts describe is an organisation whose instinctive answer, when a dependency becomes uncomfortable, is to modify copyrighted source code it has no licence to touch, switch the system back on in defiance of a vendor cut-off, and then pursue a shadow replacement while denying that the original agreement existed at all. That is not innovation. It is improvisation under legal duress, dressed up as a roadmap.

Real innovation in state IT, in the current moment, would look like things SITA has never publicly proposed. A classification doctrine that distinguishes commodity IT from sovereign infrastructure. A standing escrow regime for any system that crosses the critical threshold. A formal sunset protocol with legal cover for the official who has to sign it. A transition fund. A lifecycle owner. None of these are technically hard. They are institutionally hard, because they require the organisation to stop thinking of itself as a buyer. “Thought leadership,” as a published value, has produced none of them. What it has produced is the word “modernisation” deployed as a negotiating tactic against the vendors of systems the state can no longer operate without.

Gauteng High Court judgments have made explicit findings of mala fides, bad faith, against the state's conduct in IT disputes. That is not a procedural irregularity finding. It is not “the paperwork was wrong.” It is a judge, looking at the record, and saying that over a period of years the state denied arrangements it knew existed, used software it knew it had no lawful right to run, and reactivated systems that had been placed under interdict. Integrity, in the sense the About page uses the word, is a culture-building aspiration. Integrity, in the sense the court uses the word, is a finding about what actually happened.

A values statement and a court judgment cannot both be true in the same way at the same time. The honest thing to do, if SITA believed its own published value, would be to treat those findings as a constitutional event and rewrite the organisation around preventing them. Instead, the more characteristic response has been to argue that each one was an unusual case, an unusual vendor, an unusual set of circumstances. They are not. They are the cases in which a vendor had the resources to litigate far enough for the pattern to become visible on the public record.

eNaTIS has been waiting for a serious replacement conversation since 2007. HANIS has been inside a replacement process whose end date keeps receding since the early 2010s. The PRASA signalling estate degraded visibly for the better part of a decade while the contracts to replace it were fought over. Whatever word one uses to describe an institution whose replacement cycles are measured in decades, “agile” is not it. Neither is “adaptive.” Neither is “responsive.”

What SITA exhibits, in the judicial record, is something closer to the opposite of agility: a reflexive retreat, in every crisis, to the safest possible bureaucratic posture. “Awaiting instructions.” “Procurement irregularity under review.” “The matter is the subject of legal proceedings.” These phrases are not agility. They are the organisational equivalent of freezing. The evolving environment the value statement gestures at is the environment in which embedded systems become national infrastructure within a few years of installation, and no state IT agency in the world can afford to meet that timescale with a decade-long replacement conversation. Relevance and competitive advantage, in that environment, accrue to the institutions willing to own what they run. SITA is not yet one of them.

The collaboration value, read against the court record, is the one that stings. Judgments describe situations in which the state was actively running a proprietary system under a contractual arrangement it was simultaneously denying in writing, while in parallel commissioning work on a replacement intended to displace the same vendor whose software was still keeping the lights on. That is not collaboration. It is, in the plainest possible terms, the opposite of collaboration conducted under the heading of collaboration.

The difficult question the value raises is not whether SITA collaborates with its staff, with its board, or with other departments. It plainly does. The question is whether it can collaborate with the outside parties whose work has become infrastructure for the state, once the commercial relationship becomes inconvenient. The answer, on the record, is that it cannot. And because it cannot, the eventual resolution of every one of these disputes has to arrive through litigation, through Treasury intervention, or through collapse. None of those are collaborative instruments. They are the instruments that activate when collaboration has already failed.

This is the sentence that most completely gives the game away. Read it carefully. ICT investment must be equated with growth, impact and citizen satisfaction. The unit of analysis is investment: the act of buying. Not operation. Not ownership. Not continuity. The implicit theory of change is that the state procures the right things, and the rest takes care of itself. That is a theory that can only survive in an organisation that has not yet noticed that the hardest part of state IT is what happens after the money has been spent.

If ICT is genuinely a socio-economic game-changer, then the institution responsible for it cannot also treat it as a procurement line item. Those two framings are mutually exclusive. Infrastructure is not purchased once and then managed through a series of vendor contracts; it is inhabited, maintained, evolved, and occasionally rebuilt from the inside by people who understand it well enough to know when to stop paying a vendor and when not to. SITA's own framing, in this one sentence, collapses the category the reform depends on. You cannot be the steward of something you still describe as a purchase. And until the About page is rewritten to describe a different kind of institution, every reform proposal (including the eight in the section below) will land on an organisation that does not yet have the vocabulary to receive them.