Should AI Be Reading Athletes' Minds?

AI reading athletes' minds may seem like the start of a new fictional sports series, however, according to The Guardian, this is just the next progression for football in 2025.

Some clubs, including Bayern Munich and Brighton & Hove Albion have begun looking towards AI tools and initiatives to analyse match footage and sports' nonverbal cues. This has plenty of evidence backing with algorithms created by analysing thousands of hours of match footage and has been developed with the aim of revealing traits like confidence, leadership and even resilience. In the hope it will take their clubs a step ahead of their competitors when it comes to recruitment, enabling them to find the 'right' type of player, and root out any potential 'problem' players.

On paper this may look like an attractive proposition, being able to achieve psychological insight without the need for conversation, context, or human interpretation. It's an unconventional but potentially brilliant way of remaining ahead of other clubs and ensuring there is less risk on investment. However, in practice I can't help but feel this may become slightly murky.

Football and broadly sport in general has often struggled to quantify the mental side of performance, something many, including myself would argue does not need to be quantified. However, with football being a results based sport there is a desire for psychology to fit in with the allure of objectivity that other disciplines provide, distance covered, xG, pass completion are all key statistics which represent player performance in an easy to understand quantified way. AI offers a solution to the unquantifiable nature of psychology and human interaction by coding micro-behaviours such as posture, and gestures claiming to reveal a player's personality.

I can see how this resonates with clubs seeking a competitive edge in the market. As Max Pelka, former Bayern psychologist, notes, psychology rarely “puts numbers on the table.” AI changes that. Suddenly, a defender’s composure can be benchmarked against league peers, and a striker’s leadership inferred from a consoling pat on the back. Yet this shift toward quantification risks reducing psychology to a set of observable behaviours, stripped of context and complexity. It creates a tension between surface and depth, between what is seen and what is felt.

One of the critiques from a psychological perspective is the assumption that behaviour directly reflects internal states. A player who gestures frequently may be labeled a leader; one who avoids eye contact might be deemed withdrawn. But these interpretations are built from years of cultural, situational, and individual variability.

Nonverbal behaviour is not a universal language. A consoling gesture may signal empathy, or simply be built on habit. A lack of visible emotion may reflect focus, not detachment. Without understanding the player’s intentions, history, and interpersonal dynamics, these behaviors risk being misread.

Moreover, psychological traits are not static. Confidence fluctuates. Leadership emerges in different forms. Emotional control may look different under pressure, in training, or in private. AI, by design, captures snapshots, not stories.

There are also significant ethical implications of AI profiling. Players may be evaluated, ranked, and even selected/ traded based on data they did not consent to generate. While clubs may argue that match footage is public, the psychological interpretation of that footage ventures into personal territory. What happens when a player is told they rank in the 10th percentile for emotional control? Does it become a self-fulfilling prophecy? Does it affect their confidence, their relationships, their career trajectory.

Psychological profiling should be a collaborative process, grounded in trust and dialogue. AI risks turning it into surveillance that judges without context or conversation.

This is why the role of a human practitioner is so valuable, whilst AI can provide patterns and possibilities it cannot replace the nuanced work of a psychologist. They/ we do not just observe, we listen, question and collaboratively look for development opportunities with the knowledge of the athlete.

In practice psychological support involves navigating the ambiguity of each individual, what are their values? their experiences? and asking how does that impact their current state? Instead of labelling traits we look to understand and explore experiences rather than highlight trends.

Furthermore, psychological development is not just about identifying deficits. It’s about cultivating strengths, building resilience, and supporting identity. These processes require empathy, adaptability, and ethical care, qualities that no algorithm can accurately replicate.

Despite these concerns, AI profiling does not need to be dismissed outright. From a pragmatic perspective, it can serve as a supplementary tool to prompt reflection, support dialogue, and enhance awareness of trends and correlations. Used ethically and collaboratively with the athlete, it may help psychologists and coaches identify patterns worth exploring. However, this is with the caveat that this is a starting point not a final verdict.

As sport psychology evolves, I feel we should resist the temptation to reduce the mental side of the game into metrics. Psychology is not just about performance, it’s about people. It’s about meaning, relationships, and growth. AI can offer insight, but it cannot replace the human connection and understanding at the heart of true psychological practice.