"Training should work with physiology, not ignore it."
Artificial intelligence is becoming a training partner for millions of athletes — but much of the data behind it still reflects male physiology.
AI-powered coaching platforms promise personalised training programmes tailored to individual athletes. By analysing metrics like heart rate, training load and recovery markers, algorithms can adapt workouts in real time.
But there is a growing concern within sports science: the data these systems rely on is heavily male-dominated.
Historically, most exercise physiology research focused on male participants. Even today, female-only studies represent a small fraction of the literature.
For performance physiologist Dr Stacy Sims, this gap has long been a problem.
"Women are not small men," she said.
Her work has highlighted how hormonal fluctuations across the menstrual cycle can influence energy metabolism, temperature regulation and recovery.
Ignoring those factors risks misinterpreting female performance data.
AI systems trained primarily on male datasets may recommend inappropriate training intensities or misread recovery signals in female athletes.
In recent years, elite teams have begun addressing this gap by tracking menstrual cycles alongside traditional performance metrics. Some clubs now integrate hormonal tracking apps into their athlete monitoring systems.
The goal is not to limit training but to improve its precision.
As AI becomes more embedded in coaching, ensuring that female physiology is properly represented may become one of the biggest performance challenges in sports technology.
.png)
