AI can correct (or repeat) the relative age effect
In age-group categories, a player born in January can be almost twelve months older than another born in December. Up to a certain age, that shows in strength and speed. This is the so-called relative age effect: many youth national teams have more players born in the first months of the year.
If an AI model is trained on historical data about who was selected or promoted, it can learn that pattern and repeat it: it will recommend the most developed players and overlook those who have not yet grown. A human bias becomes an automatic bias.
AI can also help correct it. If the data includes the birth month or an estimate of physical maturity, a model can compare each player with others at the same stage of development, not just of the same year.
The key is to design systems with this problem in mind and check their results: if all the recommended players were born in the same months, something is wrong.