AI today seems to be everywhere. From science labs and hospitals to business meetings and customer service chats, it’s quietly running in the background, helping people make decisions or simply saving time. What once looked like science fiction is now part of daily life.
And you bump into it more often than you think, even when it comes to hobbies. Say you like to spend a few spare minutes in an online casino, spinning the reels or chatting with live dealers. On platforms such as Astropay casinos, which accept e-wallets like Astropay or more traditional bank cards, AI is already at work. It might be an automated system keeping your payments safe, or a chatbot guiding you through a question.
If you’re into football, you’ll see it there too—switch on a Sunday match and you won’t be surprised when AI-driven 3D models show you whether a striker was offside.
Offside calls are some of the most debated moments in football. To ease the pressure, semi-automated offside technology (SOAT) was introduced in the Premier League during matchweek 32 of the 2024/25 season. Cameras track the players, while advanced software pinpoints the exact instant the ball is played. The system then alerts VAR to a possible offside. Humans still step in to confirm the decision, but the expectation is clear—AI could one day go fully automated, making the final call itself. Whether that’s realistic or even desirable is another question.
The Subjectivity Problem
One of the biggest issues surrounding offside decisions in the beautiful game is subjectivity, and it is why the VAR team has to put eyeballs on SOAT calls. While the technology can produce extremely accurate tracking results to determine a “kick-point” (in other words, the exact moment when a ball is played forward), it can’t consider the subjectivity of the situation.
The grey areas in such subjective situations surround things like accurately identifying the right body part that’s been called offside, the relevant player and whether a player is interfering with play or not. Those small nuances are something that the system is not going to pick up on, which is why the VAR team bears the brunt of decisions.
Programming the solution
AI technology is rapidly improving all the time, and so, it’s not a stretch to imagine that it will at least get tested at some point in the future, in the hope of it being trusted to take over more and more decisions. The goal would be to orchestrate much quicker, more accurate offside decisions, but is that feasible?
Any advancement in this area would come from better programming of AI to have it be the sole arbiter of offside calls, immediately relaying its decision to the on-field referee instead of the often ponderous decision-making time that VAR teams take.
To reach that point, there would likely need to be far more sophisticated player tracking, with more cameras pointed at the game, more sensors installed and at the core level of the technology, stronger, more robust AI algorithms that the system could check itself against. That costs a lot of money, and something that’s not feasible for lower-tiered competitions to have.
The Bigger Role of Assistant Referees
Assistant referees still make the bulk of calls when it comes to offside decisions, but of course, they aren’t just there solely for that. They also help the referee out when it comes to fouls, throw-in decisions, corner and goal kicks. They provide guidance on where a throw-in should be taken from, which team gets possession for a restart, watching substitutions, ensuring that players leave and enter the field of play correctly and other tasks. Those wider observational skills of an assistant referee, arguably, can’t be replicated by an automated system.
While a specific offside detection system could probably be implemented, it would take just that one element out of the hands of officials. But that leads on to a bigger question, because if a fully automated system was found to have made an inaccurate decision retrospectively, who then is to blame? Is it the system? The developers? The league for implementing it?
Fans would have a much harder time trusting an anonymous piece of technology and would probably be far less accepting of it than if a human simply got the decision wrong.
Final Thoughts
So, could AI one day call offsides on its own? Technically, yes — and the research is already there to back it up. Some studies suggest that hybrid models combining computer vision, motion tracking and AI algorithms can flag offsides on video with an impressive accuracy rate of around 99%.
The catch is that most of this testing has been done with controlled or “canned” video, not the messy, unpredictable reality of live football — where awkward angles, crowded penalty areas and human judgement calls all come into play. That means these models aren’t quite ready to hand out final decisions without a human safety net.
In short: full AI-driven offside detection is possible down the line, and progress is moving quickly. But for now, the game still needs a referee’s eye — because the hardest part isn’t drawing a line, it’s interpreting the chaos around it.
