Artificial intelligence is no longer a promising horizon in sport: it is a constant presence, sometimes invisible, at other times uncomfortably evident. If 2025 has served to normalise its use, the question that is now starting to emerge forcefully is another: what will become of artificial intelligence in sport in 2026? How far will it really go and what will truly change?

During 2025, artificial intelligence stopped being a novelty and became a tool. There was no single foundational moment or headline that explained everything, but rather a series of small yet decisive advances. More accurate algorithms in performance analysis, increasingly refined predictive systems to prevent injuries, near-total automation in the generation of data, statistics and content. AI did not change sport, but it did change methodologies, how it is observed, measured and explained.

In many cases, the leap was not technological but cultural. Coaches, physical trainers, federations and media stopped asking whether they should use it and began to discuss how to do so without losing human judgement. Artificial intelligence began to coexist with intuition, not replace it (in the best of cases).

The limit is not technical, it is ethical

If anything has become clear after 2025, it is that the main brake on artificial intelligence in sport is not its capacity, but its legitimacy. To what extent is it acceptable for an algorithm to influence sporting decisions? Where does assistance end and unfair competitive advantage begin? What data can be used and which cannot?

The debate is no longer about whether AI can predict a performance peak or a possible injury, but about who controls that information, because there will always be clubs with more resources and federations with greater access to technology, leading to leagues that are more unequal at their starting point. Artificial intelligence threatens to amplify gaps that sport has spent decades trying to close thanks to the human component.

2026, the year of decisions

If 2025 was the year of adoption, 2026 is shaping up to be the year of structural decisions. Regulation, protocols, clear limits. Not so much to curb innovation, but to prevent it from overflowing. Sport, by definition, needs common rules to be recognisable as such and must never lose that human factor that connects it with society and with emotion (and if not, ask Formula 1, when in the past decade it became a predictable engineers’ sport with the human factor pushed aside).

In 2026 we will see to what extent artificial intelligence enters areas that have until now been protected: real-time refereeing beyond VAR, talent identification based almost exclusively on predictive models, calendar planning optimised for audiences rather than athletes. This is not science fiction. It is the next step given how this issue is progressing.

Revolution or dependence?

The big unknown is not what artificial intelligence can do, but what we want it to do for sport. It can improve athletes’ health, extend careers, reduce refereeing errors and enrich the fan experience. But it can also homogenise styles, turn the game into an excessively calculated product and reduce the margin for the unexpected (as happened to Formula 1… and which it is now trying to correct with new regulations).

The risk is not that AI gets it wrong, but that it is too right. That sport stops being a space of uncertainty and becomes a succession of well-executed probabilities.

In the end, artificial intelligence acts as a mirror of contemporary sport. It reflects its obsession with performance, its dependence on data, its ambition to grow and control every possible variable. The question is not whether artificial intelligence will go further, because it will. The question is whether sport will know how to decide how far to let it in without losing what makes it unpredictable, human and, precisely because of that, unrepeatable and passionate