Gymnastics has long relied on human eyes to separate brilliance from mistakes, but that tradition could be approaching a major turning point. World Gymnastics President Morinari Watanabe now envisions a future where artificial intelligence takes over scoring duties entirely, leaving judges with a very different role in the sport. His ambition is bold, but it also raises a bigger question: can technology ever judge gymnastics as accurately as the people who understand it?

At the World Artistic Gymnastics Championships in Rotterdam, officials overseeing all 10 apparatus had access to Fujitsu’s Judging Support System (JSS) for the first time. The technology uses cameras and AI to analyze routines, reconstruct movements in three dimensions, and detect technical details that the human eye might miss. World Gymnastics is also in advanced talks to introduce AI-assisted judging at the 2028 Los Angeles Olympics. But Watanabe wants to go further than simply helping officials. Speaking to The Athletic, he outlined his ultimate ambition: letting computers award scores without human judges.

“Our dream is to have no judges, only computers and AI awarding scores,” he said. “We must be fair to everybody. I don’t want mistakes or misjudgments.” He acknowledged that the proposal might not go down well with the officials currently responsible for scoring. Joking that they would “k-ll” him if he removed them, Watanabe accepted that humans would still need to establish rules and handle exceptional situations. For now, JSS remains a support tool, with judges retaining responsibility for the official scores.

But can AI put an end to gymnastics’ scoring controversies? Watanabe’s ambition comes against the backdrop of two major judging controversies. At the 2004 Athens Olympics, South Korea’s Yang Tae-young received an incorrect starting value on the parallel bars, affecting the men’s all-around medal standings. Two decades later, a dispute over the timing of an inquiry into Jordan Chiles’ floor score at Paris 2024 triggered a lengthy legal battle over the bronze medal.

Both incidents kept attention on judging long after the performances had ended. Watanabe believes technology could make scoring easier to understand. “Gymnastics is a wonderful sport,” he told The Athletic. “But when spectators see a score, they don’t always understand how the judges arrived at it. That’s a big problem.”

A demonstration in Rotterdam offered a glimpse of how the system works. The Athletic selected Skye Blakely’s routine for analysis, displaying her performance alongside a three-dimensional reconstruction that viewers could pause, rotate, and examine from different angles. Four cameras cover most apparatus, while floor exercise uses eight because gymnasts move across a larger area. The software identifies skills, measures joint angles, and helps explain difficulty scores, producing its analysis within 10 to 15 seconds.

Such precision could bring several benefits. AI can apply technical measurements consistently, reduce differences between judging panels, and help spectators understand why one routine receives a higher score than another. Watanabe also believes visualizing scores could make broadcasts more engaging and shorten competitions. Eventually, he hopes the technology will help coaches identify weaknesses and give athletes access to better training guidance, regardless of their location or resources.

The risks of letting computers take over

However, greater precision does not automatically guarantee better judging. Willem Standaert, an associate professor at the University of Liège who co-authored a study on AI-assisted gymnastics judging, warned that algorithms can create new problems while attempting to solve existing ones. A judge might consider a leg straight, while a computer measures it at 179 degrees. That level of exactness, he argued, may not always reflect how people understand a performance.

The technology could also struggle with skills it has not been trained to recognize. Standaert warned that completely replacing judges might weaken the sport’s development pathway, where experienced officials train and mentor the next generation. Austrian judge Johanna Gratt, who was involved in the Rotterdam event, said the system could help settle difficult decisions but still needed refinement and occasional human correction.

Accuracy and fairness raise further questions. A recent comparison with laboratory motion-capture equipment found that Fujitsu’s system measured maximum split angles an average of 1.1 degrees higher across split leaps performed by nine female collegiate gymnasts. Fujitsu maintains that its calculations follow established gymnastics rules and that its data includes athletes with different physiques, skin tones, and uniforms. However, researchers have cautioned that algorithms could perform differently when assessing bodies unlike those represented in their training data.

There is also an artistic dimension that technology may struggle to capture. Women’s floor exercise rewards choreography, musical interpretation, and expression alongside technical execution. A computer can identify a bent knee, but appreciating the artistry of a routine is another challenge.

Watanabe sees AI as a way to make gymnastics fairer and more transparent. Yet Rotterdam has exposed the difference between assisting judges and replacing them. As the sport looks towards Los Angeles, the bigger question is whether technology can understand not just the mechanics of a routine but also the human artistry that makes gymnastics special.