Olympian vs. AI: Can Technology Solve the Subjectivity Crisis in Judging Sports?

The gymnast sticks her landing. The crowd erupts. Then the scoreboard flashes a number that seems low. The arena murmurs in confusion. The cameras cut to the head judge, whose expression reveals nothing. On social media, the accusations begin: Robbed. Corrupt. Inconsistent.

It's a scene that has played out at every Olympic Games for decades. But at Milano Cortina 2026, something changed. Behind the scenes, an artificial intelligence system was watching every tumbling pass, every ice skate rotation, every boxing punch – and its silent calculations are about to upend a century of tradition.

From figure skating's AI-assisted judging debut in Milan to gymnastics' upcoming experiment in Los Angeles, the question is no longer if AI will judge the Olympics, but how quickly – and whether human judges will become obsolete.

THE PROBLEM AI IS TRYING TO SOLVE

The subjectivity crisis in Olympic judging is not new. It has names, scandals, and decades of bitter debate.

1972 Men's Basketball Final: The Soviet Union was given three chances to inbound the ball against the USA, winning a gold medal that American players still refuse to accept. A human error, not AI's domain, but it exposed the fragility of human officiating.

2002 Salt Lake City Figure Skating Scandal: French judge Marie-Reine Le Gougne admitted she was pressured to vote for the Russian pair over the Canadians. The resulting outcry forced the IOC to overhaul judging systems, introducing anonymous scoring and randomized judging panels. But the problem – human bias – remained.

2016 Rio Boxing: Multiple refereeing decisions were so controversial that the IOC suspended all referees and judges involved, eventually stripping the AIBA (now IBA) of its right to organize Olympic boxing for Tokyo 2020.

Tokyo 2020 Gymnastics: Italian gymnast Vanessa Ferrari's floor exercise score caused an uproar when her difficulty rating seemed miscalculated. Russia's Nikita Nagornyy openly complained that judging varied wildly between apparatuses.

At the heart of every controversy lies the same uncomfortable truth: human perception is flawed, fatigue is real, and bias – conscious or unconscious – is unavoidable. Even with multiple judges, redundant scoring, and instant replay, the margin for error remains.

Enter artificial intelligence.

HOW AI JUDGING ACTUALLY WORKS

AI judging systems have been quietly tested for years. The technology falls into three categories.

Computer vision scoring. Cameras capture an athlete's performance from multiple angles. Machine learning models, trained on thousands of previously judged routines, identify key technical elements: body angles, rotation speed, landing positions, foot alignment. The AI then assigns a "raw score" based purely on objective measurements – no politics, no fatigue, no favoritism.

Real-time violation detection. In sports like boxing or taekwondo, AI can track whether a punch lands cleanly (above the belt, closed fist, proper technique) or whether a kick strikes the legal target area. Systems have already been tested in professional boxing exhibitions, where sensors in gloves and body armor feed data to an AI that determines scoring blows faster than any ringside judge.

Predictive and prescriptive analytics. More advanced systems don't just score what they see – they compare performances to historical data to identify anomalies. For example, an AI might flag that a gymnast's vault rotation was 15 degrees off vertical, automatically deducting the appropriate points, or note that a figure skater's combination jump had a slight pre-rotation that human judges often miss.

The standard bearer for AI judging development is the Fusion System, a platform developed by Fujitsu and the International Gymnastics Federation (FIG). Fujitsu has been refining the technology since 2017, using skeleton modeling and 3D laser sensors to digitize athlete movements with precision down to fractions of a second.

"We can visualize the invisible," said a Fujitsu engineer at a 2025 demonstration. "The human eye cannot track a gymnast's hip angle during a triple twist. The AI can, and we can show the judges exactly what we see."

THE MILANO CORTINA 2026 TEST RUN

At Milano Cortina, AI judging made its official Olympic debut – but in a limited, cautious form. The figure skating short program used an "AI-assisted review system" that flagged potential under-rotations and edge calls for human judges to review. The AI didn't change any scores automatically. Instead, it acted as a second pair of eyes, highlighting moments that might have been missed in real time.

The results were mixed.

In the men's singles, AI flagged two separate jumping passes that human judges had scored as clean. Replay confirmed the under-rotations. The athletes' scores were adjusted downward with no subsequent protest – the evidence was indisputable.

In women's singles, an AI alert led to an extended review that delayed medal ceremonies by nearly two hours. Both the skater and her coach argued that the AI had misidentified a slight blade scratch as an under-rotation. The human judges overruled the AI, but the damage was done: fans left the arena confused, and social media raged about "robot refs."

One figure skating judge, speaking anonymously to a French sports daily, admitted, "The AI saw things we never would have caught. But sometimes, the things it caught didn't actually matter to the quality of the performance. The system doesn't understand artistry – only geometry."

That tension – between technical precision and artistic expression – is the central conflict of AI judging.

THE ARTISTRY PROBLEM

No AI can yet judge artistry. It can measure a split leap's angle, count the milliseconds a balance is held, or detect a flubbed landing. But it cannot feel the emotional resonance of a floor routine set to haunting music, or appreciate the narrative arc of a free skate that tells a story.

This is not a minor limitation. In sports like figure skating, synchronized swimming (now "artistic swimming"), rhythmic gymnastics, and even freestyle skiing's moguls and aerials, performance is judged as much on presentation, musicality, and creativity as on pure technical execution. AI, at least for now, is blind to those dimensions.

"I would never want a machine to judge my expression," said a three-time Olympic figure skater in a 2025 interview. "The whole point of our sport is to make people feel something. A camera doesn't feel. A sensor doesn't weep. If we let AI take over, we might as well be robots ourselves."

The International Skating Union (ISU) has acknowledged the problem. Their current approach – used in Milano Cortina and likely continued for LA28 – is a hybrid model: AI handles technical elements (rotations, edges, under-rotations), while human judges score artistry, interpretation, and overall impression.

But hybrids come with their own issues. What happens when the AI's technical score suggests a performance should be gold-medal worthy, but the human artistry score pulls it down to bronze? Who wins the tiebreak? And do fans trust a system that splits authority between humans and machines?

BOXING, GYMNASTICS, AND THE COMING WAVE

Figure skating's experiment is just the beginning. Several sports are preparing AI expansions for LA28 and French Alps 2030.

Boxing: The sport is the most urgent case. After the IOC stripped recognition from the IBA in 2023 (due to governance and financial issues), boxing's Olympic future was uncertain. A new body, World Boxing, has emerged, and part of its pitch to the IOC is technology-driven judging. For LA28, World Boxing expects to deploy sensor-embedded gloves and vests that automatically register clean punches, scoring them in real time. Ringside judges would still oversee the fight, but their role would shift from scoring every exchange to overriding the AI only in ambiguous cases.

Gymnastics: FIG and Fujitsu are already planning a more advanced version of the Fusion System for LA28. The goal is full AI scoring for vault (where technical elements are most discrete) by 2028, with other apparatuses to follow by 2032. Floor exercise and balance beam will likely retain human judges for artistry longer than vault or parallel bars.

Snowboarding halfpipe (Winter): French Alps 2030 is expected to introduce AI-assisted scoring for snowboard halfpipe, where rotation counts and landing accuracy are relatively straightforward for computer vision. Human judges will still score "amplitude" (height of tricks) and "flow" (variety and creativity).

Taekwondo: Electronic scoring for kicks has existed for years (using sensors in body armor). The next frontier is AI detection of illegal strikes (to the head, below the belt) and automatic penalty assessment – removing the referee's subjective judgment from foul calls.

WHAT ATHLETES THINK: THE GREAT DIVIDE

Supporters point to consistency. "Human judges have favorites," said a former Olympic gymnast. "They know who the star is before you even walk out. AI doesn't care if you're Simone Biles or a first-time Olympian. It just measures your body."

Critics argue AI strips sports of their soul. "We are artists and athletes," said a figure skater. "If you only measure my rotations, you miss the whole point. The Olympics are about humanity, not engineering."

A 2025 survey of Olympic medalists (conducted by the IOC Athletes' Commission) found that 61% of athletes in judged sports opposed fully autonomous AI judging, but 78% supported AI-assisted review systems similar to Milano Cortina's model. The consensus: AI as a tool, not a replacement.

THE LA28 EXPERIMENT AND BEYOND

For the Los Angeles 2028 Summer Games, the IOC plans the most ambitious AI judging deployment yet. Gymnastics (vault), taekwondo, and boxing are expected to use sensor-based or computer-vision systems for at least preliminary scoring. Figure skating and artistic swimming will likely remain hybrid, with humans deciding artistry while AI flags technical errors.

But the ultimate test – full AI autonomy, with no human judges – is still years away, at least for sports with an artistry component. The first discipline to go fully automated may be weightlifting (where sensors can measure bar path and lockout position) or archery (where lasers already determine arrow placement). Neither sport involves subjective expression, making them ideal candidates.

THE PHILOSOPHICAL QUESTION

Beyond the technology, AI judging forces a deeper question: What are the Olympics for?

If the Olympics are purely a test of objective athletic performance – who jumped highest, ran fastest, threw farthest – then AI should eventually judge everything. Machines are better than humans at measuring speed, distance, rotation, and time.

But if the Olympics are also a celebration of human spirit, will, and artistry – if we watch figure skating not just to count rotations but to feel something – then human judges will always have a role. And they will always make mistakes. And those mistakes will always infuriate fans.

The future of AI judging, then, is not a technical problem. It is a choice about what kind of Games we want.

For now, the IOC is choosing caution. Milano Cortina was a test. LA28 will be another test. And somewhere down the line, perhaps in Brisbane 2032 or Salt Lake City 2034, a computer will score an entire Olympic final without a single human judge's input.

When that day comes, we will know exactly how much we value precision – and how much we are willing to trade for a perfect score that nobody cried over.

This is independent Olympic news coverage. For sports nutrition and performance supplements, visit dadssupplements.com.