AI in Sport - Opportunity, Excitement, Progression and Acceleration

‍ARTIFICIAL INTELLIGENCE IN SPORT

Artificial Intelligence (AI) will revolutionise all aspects of sport -- enhancing athlete performance, improving on pitch decision-making, and elevating fan experiences. But too often, its real-world influence gets drowned out by overblown hype. 

To help rectify this, Vice-Chancellor and world leading researcher in AI from my own Loughborough University, Professor Nick Jennings, has joined with me to write this piece in which we examine some of the most profound short to medium term impacts of this exciting technology.

The Power of AI in Sport

There are numerous examples of the impact of AI in sport, but here we focus on a few of the most important -- identifying talented athletes, accessing top quality coaching, turbocharging anti-doping testing and enhancing fan experience.

AI can be a great leveller and can help developing nations and those with less mature sport ecosystems to discover hidden talent and overcome geographic and financial barriers. Helping develop these robust and powerful uses of AI is the only way to unlock this true potential.

Identifying Talented Athletes

AI is transforming the way athletes are scouted and developed. Traditional talent identification processes rely on subjective assessments and limited data, which result in missed opportunities. AI, on the other hand, can analyse vast amounts of data from various sources, such as performance metrics, physiological data, and even social media activity, to identify promising athletes at an early age. 

By exploiting video processing and machine-learning algorithms, AI can detect promising individuals from uploaded mobile phone videos which significantly widens the talent aperture. From such inputs, the algorithms can highlight those who are worthy of further investigation and so focus the efforts of human scouts onto those with the most promise. 

Having identified talented individuals, AI-powered platform scan analyse footage of athletes in action over time, assessing their technique, speed, agility and other performance indicators. This data can then be compared against benchmarks and historical data to assess the trajectory of exceptional talent. In so doing, AI can continuously monitor an athlete's progress, providing real-time feedback and recommendations for improvement. This proactive approach ensures that talented athletes are identified and nurtured from a young age, maximising their chances of success.

A great example of this is the NFL Digital Athlete - an innovative initiative aimed at enhancing player health and safety through the use of AI and machine learning. This collaboration between the NFL and AmazonWeb Services, collects and analyses data from various sources, including game footage, player tracking systems and sensors embedded in equipment. By creating a comprehensive view of players' experiences, the Digital Athlete helps teams understand individual needs, predict and prevent injuries and develop personalised training and recovery programmes. 

‍

Accessing High Quality Coaching

AI is revolutionising the way athletes access coaching and training resources. In many regions, particularly in developing nations, access to high-quality coaching is limited due to geographic and financial constraints. AI can bridge this gap by providing personalised coaching programmes and virtual training sessions. AI-driven coaching platforms can analyse an athlete's performance data and generate customised training plans tailored to their specific needs and goals.

For example, Dinetiq (a Loughborough spin out) offers a range of data-driven, science-backed coaching services for cricket fast bowlers. Their system embeds 20 years of world-leading research into a mobile phone app that provides personalised advice to individuals to improve their performance and reduce injury risk.

AI can also simulate game scenarios, allowing athletes to practise and refine their skills in a controlled environment. Virtual coaching sessions can be conducted through video conferencing or virtual reality platforms, enabling athletes to receive real-time feedback and guidance from expert coaches, regardless of their location. In so doing, AI systems can provide athletes with access to a wealth of training resources, such as instructional videos, drills, and exercises, which can be accessed at any time. This democratisation of coaching ensures that athletes from all backgrounds and all skill levels have the opportunity to improve and reach their full potential.

‍

Turbocharging Anti-Doping Testing

AI is playing a crucial role in enhancing anti-doping efforts by improving the detection and prevention of doping violations. Traditional methods often rely on random testing and manual analysis, which can be time-consuming and prone to human error. In contrast, AI systems can analyse large datasets with much greater frequency and can, therefore, more effectively identify anomalies that may indicate doping. 

In particular, machine-learning algorithms can be trained to recognise patterns associated with doping, such as sudden changes in an athlete's performance metrics or physiological data. By continuously monitoring this data through trusted measurement sensors and devices, AI can flag suspicious activities and trigger targeted testing, increasing the likelihood of detecting doping violations. Additionally, AI can analyse data from various sources, such as biological passports, to identify long-term trends and detect subtle changes that may indicate doping. This proactive approach not only enhances the effectiveness of anti-doping efforts but also acts as a deterrent, discouraging athletes from engaging in doping practices.

Enhancing Fan Engagement

AI can significantly enhance fan engagement by offering personalised experiences tailored to individual preferences. By analysing fan behaviour data—such as viewing history, preferences, and social media activity—AI systems can deliver customised content. This can include match highlights, news, and product recommendations, all designed to engage fans more deeply by providing content that aligns with their specific interests.

Additionally, AI-powered chatbots and virtual assistants are transforming fan interaction. These technologies can handle fan queries, provide real-time game updates, and foster engagement during events via social media platforms. By maintaining an active presence beyond the event itself, these tools enhance the overall fan experience, keeping fans connected and engaged with the sport in real time.

Furthermore, AI can power virtual and augmented reality(VR/AR) experiences that immerse fans in an interactive and dynamic viewing environment. Fans can experience the thrill of being courtside or standing on the field, creating a more engaging and memorable connection with the event. On top of this, AI-driven predictive analytics can assist fans in fantasy sports by offering data-backed insights to improve team decisions, considering factors like player performance trends and injuries.

Challenges

Ethical and privacy concerns are paramount when it comes to safeguarding athletes' rights and autonomy, particularly as AI increasingly plays a role in tracking and analysing personal data. Real-time data accuracy is essential for providing actionable insights during training and competition, where even minor inaccuracies can have a significant impact on performance and decision-making. To address these concerns, it is crucial to implement robust governance, enhanced security measures, and improved data protocols that protect athletes' rights while maximising AI’s potential in enhancing sport performance and decision-making.

The ethical and privacy implications of AI are especially critical in the context of sensitive data, such as athletes' health metrics, performance statistics, and biometric details. While these datasets are invaluable for optimising training and performance, they also carry significant risks of unauthorised access, misuse, and discrimination. Without proper regulation, AI applications could lead to invasive monitoring or data breaches. Fortunately, modern privacy preserving technologies can now store, manipulate and aggregate such data (using approaches like aggregated conversion modeling and homomorphic encryption) in a safe way without it being shared or distributed widely.   All of this is paramount but still does not address the sadly all too familiar use of AI in online abuse and cyber bullying.  It is rare for this type of abuse to be solely AI generated.  There is inevitably a malicious human intending to cause harm sitting behind the abuse that many of our athletes have unfortunately been targets of.  Here we need vigilance, law enforcement and social media platform leadership to track, take down and prosecute the individuals. At World Athletics we have conducted a 4 year study of online abuse of our athletes from the Tokyo 2020 Olympic Games to the Paris 2024 Games, taking in two of our own World Championships. The reports make unpleasant reading but through reporting, to platforms and police, we have had some success in removing abusive posts and handing individuals over to local law enforcement. The legal system is catching up but we need stronger partnerships with social media platforms to eliminate this scourge.

Conclusion

AI is a game-changer for sport, offering unprecedented opportunities to discover talent, enhance performance, ensure competition is fair and better engage fans. By leveraging AI's capabilities, we can unlock new possibilities, particularly in regions like Africa, where there is immense untapped potential waiting to be realised. The future of sport is bright with AI. It is time to take action and harness its power for the betterment of athletes, fans and the entire sports community.

Co-authored by Seb Coe and Prof. Nick Jennings

‍