The most influential figure at the World Cup may never touch the ball. It doesn't wear boots, it isn't on any team sheet, and most fans in the stadium will never see it. It lives in millions of lines of code — and unlike a player, it never waits for permission.
Artificial intelligence has quietly become football's newest signing. The interesting question is no longer whether it belongs in the game. That argument is over. The question is what football becomes once the machines are this deep inside it.
Editor's note: This is an opinion piece. The technologies described — semi-automated offside, the connected match ball, limb-tracking cameras, predictive analytics — are real and in use; the conclusions drawn from them are ours.
The invisible teammate
For most of its history, football sold itself on being beautifully simple. Twenty-two players, one ball, one referee, and a great deal of instinct. That version of the game is quietly disappearing.
AI now sits in places fans rarely think about. It helps clubs scout and recruit by surfacing players who fit a system before a human analyst ever watches the tape. It helps medical teams flag injury risk from workload and movement data. It helps coaches model the opposition, and it helps referees rule on the tightest margins in the game. Before a player makes a run, software has already weighed thousands of comparable movements. Before a manager picks a lineup, a model has run the probabilities. Before a referee signals a goal, a system has checked whether it should count.
Football is no longer only being played. It's being processed.
From VAR to genuine automation
The first real shock was VAR. Fans loved it and hated it, often inside the same ten minutes. But whatever you think of it, it changed the game's relationship with certainty for good.
What's arrived since is a different order of thing. Semi-automated offside technology — used at the 2022 World Cup and refined since — combines multiple tracking cameras with a sensor inside the match ball to pin down the precise moment the ball is played. Adidas's connected balls carry a motion sensor sending data hundreds of times a second, while limb-tracking cameras map up to dozens of points on every player's body. A decision that once took a VAR official minutes of squinting at frozen frames can now be reached in seconds, with an automated alert and a 3D animation to explain it.
The goal is honest enough: make football fairer and faster to adjudicate. The tension is just as obvious. Can the game get more accurate without getting less human?
The Guardiola problem
Imagine handing every coach unlimited information. That is roughly what modern football has done. Recruitment departments run predictive models on talent. Analysts pull apart an opponent's weaknesses with data the opponent can also see. Sports scientists use machine learning to anticipate the muscle strain before it happens.
In theory, smarter inputs make smarter football. In practice it raises an awkward question: if everyone is drinking from the same data, where does an edge — or originality — come from? Football's defining moments were rarely the calculated ones. No model forecast Maradona's run against England, and no algorithm would have drawn up Messi drifting through a defence in Doha. The magic tends to live in the unpredictable, and unpredictability is famously hard to code.
The referee's new assistant — or new burden
Referees may be AI's biggest winners. They may also be its biggest victims. Every tool that promises fewer mistakes also raises the bar for what counts as a mistake.
Fans once accepted human error as part of the bargain. Now that a decision can be checked frame by frame, every call is held to that standard, and the outrage when technology still gets one wrong is louder, not quieter. The strange result: the more accurate football becomes, the less tolerant the crowd gets. AI isn't killing controversy. It's just moving the target.
What comes next
The next phase is already taking shape in elite setups: tactical suggestions generated live during a match, real-time fatigue and injury alerts pushed to the bench, automated opposition reports produced within minutes of the final whistle, training plans that update after every session. What sounds futuristic now has a way of becoming standard within a season or two. Football has always chased marginal gains; AI is selling something larger, and measuring the return in trophies.
The worry nobody wants to say out loud
There's another side to all this. Football has always belonged to people — the scout who spotted a talent on a muddy pitch, the coach who trusted a hunch over the spreadsheet, the player who ignored the instruction and made something nobody planned. As AI grows more influential, the game inherits a dilemma every industry eventually meets: how much judgement should humans hand over?
Because football's greatest asset was never efficiency. It was imagination.
Offside
AI isn't coming to football. It's here, in the ball, in the cameras, in the recruitment meeting, in the medical room. The debate about whether it belongs is finished.
The more interesting question is the one we'll be answering for years: as football gets smarter, can it stay surprising? Fans don't fill stadiums to watch perfect decisions. They come to witness the moment nobody saw coming. If the game ever trades that uncertainty away for accuracy, not even the smartest machine in the world will be able to hand it back.
If you want the analytics side of this story explained plainly, start with our guide to expected goals (xG) — and for the technology Musk-style empires are racing to build, see getting started with AI agents.
