Can AI Replace Human Referees in the Future?

In the current world of sports, particularly football, the question of whether AI can substitute human referees is becoming increasingly apparent. Already with the current technology such as VAR (Video Assistant Referee) in use, AI promises more accuracy and objectivity. Technologies like VAR and semi-automated offside technology (SAOT) are offering more accuracy, eliminating human error for calls like goals and offsides.

AI is revolutionizing refereeing in sports by making effective objective decisions. For instance, tennis' Hawk-Eye uses AI for line calls, and the ATP aims to have all line judges substituted by 2025. VAR in football helps with offside and goal-line, reducing errors. A study that was delivered at Frontiers in Sports and Active Living shows AI does well when it comes to speed and accuracy for clear calls but badly in subjective fouls. Sports websites like Deseret News highlight how AI reduces human bias, something that may make games more level.

Supporters argue that AI can put an end to controversy. A Vanderbilt Journal of Entertainment and Technology Law article records AI's potential to make officiating larger games better. Teams like FC Barcelona, in their Innovation Hub, discuss referee-less football, using tennis as a precedent.

Nevertheless, the area of sports where AI can be used may fail at subjective decisions, i.e., discriminating fouls or intent, which are more competent to the human factor. John Eric Goff assert that AI is in no position to replace referees because of the lack of emotional background and contextual understanding. Spectators appreciate the human element, too, with surveys indicating ambivalence regarding VAR interruptions to the flow of the game (Hamsund & Scelles, 2021). Technical failures, such as the 2020 Premier League goal-line mistake, illustrate AI limitations.

Even though AI cannot yet completely replace the human referees because of contextual interpretation, a combination of both technologies is still eminent. The most feasible direction is a hybrid model to leave objective calls to AI and complex decisions to an unreliable entity, a human. According to a Frontiers article, the best option is to combine AI pre-reviews with human validation to achieve maximum results. The future? not AI-driven but AI-assisted.