At the end of July, many people across the globe were preparing to tune into the two-week, 2024 Olympic Games in Paris, France. The Olympics were slated to feature several high-profile athletes—including Simone Biles (USA, artistic gymnastics), Eluid Kipchoge (Kenya, marathon) and Marta Vieira da Silva (Brazil, football). However, in the lead-up to the Games, the International Olympic Committee (IOC) focused on a secondary player: Artificial Intelligence (AI). The IOC laid out an ambitious AI agenda aimed to enhance athlete performance, ensure fairness and optimize operations. The 2024 Paris Olympics represent a significant leap forward in integrating AI into the world of sports.
“Together, we can unlock AI’s full potential to promote solidarity, further digitalization, improve sustainability and resilience, and reinforce the role of sport in society.” – Olympic AI Agenda
Here, we explore several applications of AI in the 2024 Paris Olympics.
AI-Driven Athlete Training and Support
Olympic athletes spend years—even decades if you consider Diana Taurasi (USA, women’s basketball) in her sixth Olympic games—training at a high level to compete at the Olympics. AI has transformed the way athletes train, from monitoring for injuries to developing individualized training and nutrition plans.
A large portion of the IOC AI agenda emphasized using AI to train athletes and to scope out potential talent. Many coaches and athletes have implemented AI in their training programs in the days leading up to the Paris Olympics. For example, the Chinese Olympic Diving team used AI to capture diver movement where traditional video recordings fall short. Lu Feixiang, one of the developers of the AI training system at Baidu, shares that the technology can capture intricate details of diving movements and provide real-time analysis. This capability allows coaches and athletes to make slight adjustments throughout training. The results were evident as the Chinese Olympic diving team swept the competition, winning six gold medals and two silver medals. It is fair to say that implementing AI in their training regimen was highly beneficial.
AI is already ever-present and seamlessly integrated in many of our smartphone apps, enhancing our daily lives through personal assistants and personalized algorithms. The IOC sought to leverage AI in a similar way to support athletes while during their time in Paris. In the Olympic Village—where many of the athletes stay throughout the two-week competition—the athletes had access to an IOC Assistant Chatbot, called “AthleteGPT”, powered by Intel. According to Intel, AthleteGPT can could assist with navigating the venue, adhering to Olympic Village guidelines, and support over 11,000 athletes in multiple languages. All of this in real-time through voice or text commands on a smartphone application. This advancement allows athletes to focus on their training and competing throughout the Olympics.
Heighted Protection from Online Harassment
In addition to athletes’ physical safety, the IOC was committed to protecting athletes from online harassment. For two weeks, athletes are performing in high-stakes, high-pressure environments and are under immense scrutiny. Because of this, the IOC implemented an AI-monitoring system to screen social media activity during the games. The system was responsible for protecting over 15,000 athletes and 2,000 officials during the Games. While social media platforms often screen for specific words to weed out harmful content, they can struggle with the complexities and nuances of language. The AI-monitoring system harnessed Large Language Models and a “Threat Matrix” to safeguard athletes.
Large language models—like ChatGPT—learn to interpret and generate text by analyzing vast amounts of data. These models are adept at discerning the underlying emotions and intentions in a piece of text, even when obvious indicators are missing. The IOC’s system for combating online abuse, known as “Threat Matrix”, is specifically designed for this purpose.
Threat Matrix employs a comprehensive approach, as highlighted by Qublai Ali Mirza, a senior lecturer at the University of Gloucestershire with expertise in AI and cyberbullying. One key element is sentiment analysis, which uncovers the emotional tone of a message. For instance, detecting sarcasm—a task humans excel at but AI is just beginning to master—is within the capability of Threat Matrix. Additionally, advanced systems like Threat Matrix can interpret how images and emojis alter the meaning of text, all while navigating the diverse nuances of different languages and cultures. Achieving this manually would require an impractically human large team, even for an organization as well-resourced as the IOC.
Large language models are a huge stride in implementing AI to shield athletes from online harassment, but Kristy Burrows (Head of the Safe Sport Unit in the IOC) says: “Even with the power that AI affords, online vitriol is a societal problem, and technology can’t solve societal problems alone.”
Enhanced Precision in Officiating
The use of AI in judging and refereeing brought exceptional levels of accuracy and impartiality to the Olympics. AI systems can provide real-time analysis of events, reducing human bias and ensuring fair competition. AI has already made a significant impact in sports like football (soccer, if you’re in the USA), where an array of cameras around the stadium and chips implanted in the ball are seamlessly collecting data throughout match play. These technologies assist referees in making precise decisions like if a ball crossed the goal line or if a player is offsides. This implementation has significantly enhanced the fairness and reliability of officiating in football, a trend that continued throughout the Paris Olympics.
Omega, the company tasked with keeping time at the Olympics, is not using AI to decide winners and losers of races, but rather to simultaneously collect data and produce next-generation stats like strokes per second or acceleration/deceleration in real-time. Using this information, broadcast teams were able to create enhanced data graphics for viewers. A prominent example of graphics including real-time, next generation stats took place in the pool when American swimmer, Katie Ledecky, won gold in the 1500-meter swim. She was meters ahead of her competition and TV viewers saw the word “Leader” stamped in her lane on screen. The lanes of second and third place swimmers displayed dynamic metrics conveying how many seconds behind they were in catching Ledecky. Spoiler: they never caught up to Ledecky and she finished ten seconds ahead of the silver medalist, Anastasia Kirpichnikova of France.
The Future of AI in Competition
The integration of AI into the 2024 Paris Olympics has demonstrated the technology’s transformative potential across various aspects of the Games. From enhancing athlete training and safety to improving the accuracy and fairness of judging and refereeing, AI has played a crucial role in elevating the Olympic experience. These advancements have benefited the athletes and enriched the viewing experience for millions of fans around the world.
Looking ahead, the continued evolution of AI promises even greater strides in the world of sports. Future Olympic Games could see more sophisticated AI applications—like advanced injury prediction models, real-time tactical analysis for coaches and even more immersive fan engagement technologies. As AI technology becomes more accessible and refined, its potential to revolutionize sports will continue to evolve. However, as Kristy Burrows aptly noted, technology alone cannot solve all societal problems. The responsible and ethical use of AI, combined with human oversight and societal efforts, will be essential to harness its full potential. The AI race has just begun, and the future of AI in sports looks incredibly promising.
Anna Bennett
