After Dimeji Shittu’s Commonwealth Games final left commentators screaming ‘robbery’ louder than a fan who just watched their team lose on a VAR review, the boxing world has arrived at the inevitable conclusion: we need machines to make our terrible decisions.
The logic is flawless, really. Humans see a punch land clean and somehow score it for the other guy. So the answer is obviously to replace them with algorithms trained on footage of humans seeing punches land clean and scoring them for the other guy. We are simply outsourcing the confusion to a server farm in Dublin.
Let us be clear about what we are actually doing here. We are taking a sport where judges already disagree with each other, with commentators, with the fighters themselves, and with basic physics—and we are adding a third layer of incomprehension by asking a neural network to learn from all that disagreement. The AI will not magically solve boxing judging. It will simply produce wrong decisions at the speed of light, with the added bonus that nobody can explain why.
Shittu was robbed. That much is obvious to anyone with functioning eyes. But the solution is not artificial intelligence. It is actual judges who watch the fight, understand boxing, and do not have financial incentives to favour one fighter over another. Revolutionary, I know.
But here we are in 2026, convinced that every human failure can be solved by feeding it to a machine. Boxing judging does not need disruption. It needs people who know what they are watching.