Tennis has always been a sport obsessed with numbers. Aces, double faults, break points, first-serve percentages, winners and unforced errors have long been used to explain why a match turned one way or another. But at the 2026 US Open, the most important number may no longer be on the scoreboard. It could be the number generated by artificial intelligence.
IBM and the United States Tennis Association are taking the digital experience at Flushing Meadows another step forward this year, transforming the US Open from a tournament fans watch into one they can increasingly interrogate, understand and personalise. The partnership, which dates back to 1992, now serves more than 14 million global tennis fans annually through the US Open website and app.
The real significance of IBM’s AI push is not that machines can predict who will win. It is that they are beginning to explain why a match is changing. That shift from prediction to understanding could fundamentally change how fans consume live sport. The headline feature is Match Chat, an AI-powered companion that allows fans to ask questions in natural language during and after singles matches.
Instead of navigating through statistics or searching separate databases, fans can simply ask about a player’s head-to-head record, the key statistical battle, what has changed during the match or even how a player’s name is pronounced.
The 2026 version goes further. Match Chat, powered by IBM watsonx Orchestrate, can draw from live match data, historical information and analysis, with some answers now incorporating photographs and video. IBM says the underlying system uses multiple AI agents and models trained around the USTA’s editorial style and the language of tennis. That is an important development. AI in sport has often been sold as a prediction machine. Fans are shown percentages and probabilities, but percentages without context can be meaningless.
IBM is now trying to solve that problem. Its Likelihood to Win tool continues to calculate a player’s probability of victory as the match develops, updating after every point. The model considers current and historical statistics, expert opinion and match momentum. But this year’s US Open adds Key Moments, which attempts to identify the moments responsible for major shifts in the match. Knowing that a player’s win probability has jumped from 42% to 68% is interesting. Knowing that the shift happened because of a crucial break, a sequence of dominant rallies or a costly double fault is far more useful.
It turns AI from a scoreboard into an interpreter. The US Open is also introducing a new Serve Quality metric across all 254 singles matches. The technology uses limb-tracking to examine 21 data points across the player’s body and racket, tracking movement 50 times a second. IBM says this will generate roughly 1.2 billion data points across the tournament.
For the average viewer, that may sound excessive. But this is precisely where sports technology is becoming more interesting. A serve is no longer simply an ace or a fault. AI can increasingly examine the mechanics behind it and translate those movements into information that fans can actually understand.
The bigger story, however, is happening behind the screen. IBM’s relationship with the US Open has evolved over more than three decades. What began with digital infrastructure has developed into a hybrid cloud and AI ecosystem designed to handle enormous volumes of information and traffic. IBM says the tournament processes more than 1.2 billion data points, while the digital platform has to cope with traffic spikes that can exceed 5,000% during high-interest matches.
That infrastructure is important because AI is only as useful as the data feeding it. Every point at the US Open generates information about serve speed, shot placement, rally length and other aspects of performance. Turning that stream into something meaningful in seconds is the technological challenge. The fan sees a prediction or an insight. Behind it is an enormous data operation designed to make that insight appear almost instantaneously. The US Open’s AI ambitions also reflect a broader transformation across tennis.
At Wimbledon this year, IBM introduced its own enhanced Match Chat and Key Moments features. The latter was specifically designed to explain the reasons behind changes in win probability, rather than simply presenting a number. That parallel development is telling. The world’s biggest tennis tournaments are beginning to converge around the same question: how can technology make an increasingly data-heavy sport easier, rather than more complicated, to understand?
IBM’s Jonathan Adashek captured the broader ambition when he said, “As fan engagement and content consumption habits evolve, IBM and the USTA are at the cutting edge – leveraging data and AI to create unique digital experiences that bring the excitement of the US Open to life for audiences around the world.” The challenge is ensuring that AI remains a servant of the sport rather than becoming its distraction.
Tennis does not need a machine to tell fans whether a five-set thriller was exciting. It does not need an algorithm to replace the emotional intelligence of a commentator or the instinct of a spectator watching a comeback unfold. What it can use is context. Why did the momentum change? Why did a player suddenly lose control of a match? Which part of the game is deciding the contest? What happened while the fan was away from the screen?
Those are questions AI can potentially answer better than a traditional scoreboard. And that is why the 2026 US Open represents more than another showcase for IBM’s latest technology. It is an experiment in how sport itself can be consumed in the age of AI.
