FIFA is deploying AI to run the 2026 World Cup with 48 teams. From Football AI Pro to camera referees and 3D player models, the tournament will become a major test for AI in football.
When Romy Gai, FIFA's Chief Business Officer, spoke about the challenges of running a World Cup with 48 teams spread across three countries—Canada, Mexico, and the United States—he wasn't emphasizing technology. What he was referring to was the sheer complexity of the event.
In previous World Cups, local organizing committees handled the majority of the logistical work. But at the 2026 World Cup, FIFA will directly oversee the entire operation. Approximately six billion people are projected to watch the tournament. The total number of matches has increased to 104, compared to 64 in Qatar. The number of participating teams has risen from 32 to 48, with over 180 broadcasters involved, and no single national infrastructure to rely on. The scale of this tournament is truly unprecedented.
The AI strategy that FIFA announced at the Lenovo Tech World event in Hong Kong this week needs to be viewed in this context. Notable announcements include Football AI Pro, 3D AI-powered player avatars, and a next-generation version of Referee View. However, these product decisions reflect a larger structural shift: FIFA has decided that AI is not just a tool to support the World Cup, but the foundation for running the entire tournament.
Football AI Pro is an AI-generated knowledge assistant that will be provided to all 48 teams participating in the 2026 World Cup. The system is built on FIFA's Football Language Model and trained using hundreds of millions of data points owned by FIFA. It can generate pre- and post-match analyses in text, video, chart, and 3D model formats, and supports multiple languages. However, this tool will not be used during matches.
The core idea behind this system is to democratize football analytics data. At the highest level of the sport, access to modern match analysis tools is heavily dependent on each team's financial resources. Top teams often have entire data analytics departments, while teams making their World Cup debut almost never have that. Football AI Pro is designed to ensure all teams have the same fundamental analytics platform.
However, deploying this system also presents a significant challenge in terms of enterprise AI infrastructure. Providing consistent analytical data to 48 teams across three different countries, in multiple languages, and throughout a multi-week competition schedule is a considerable technical task. This is precisely the kind of workload that requires the hybrid AI architecture that Lenovo is developing for its enterprise solutions.
Another notable technology is Referee View, a camera-based perspective of the referee. The new version uses AI for real-time image stabilization, reducing the shaking that made the previous version difficult to follow during high-speed plays.
However, the primary goal of this technology is not just to serve television broadcasts, but to increase transparency. VAR technology has been controversial in football, partly because the decision-making process is difficult for viewers to understand, and partly because the illustrative images are sometimes not clear enough. Providing better quality, near-real-time camera angles from the referee can help viewers better understand decisions on the field.
The first version of Referee View was tested at last year's FIFA Club World Cup. The update for the 2026 World Cup is a significant technological leap, but the real challenge is whether it will change how fans perceive referee decisions. If it does, the technology will become a governance tool, not just a broadcasting tool.
Furthermore, the AI-powered 3D player avatar system is designed to address a long-standing problem in football: semi-automatic offside detection technology. The current system works well, but the images used to explain offside decisions are sometimes difficult to understand. The lines are hard to read, the viewing angles are not intuitive, and fans often argue even when the technology's decisions are correct.
The new system will scan players to create accurate 3D models, with each scan taking only about a second. During a match, these models help track player movement more accurately, even during collisions or when the view is obstructed.
When an offside situation is brought up by VAR, the 3D model will create a more accurate and easier-to-understand illustration for viewers. This technology was tested at last year's FIFA Intercontinental Cup, when Flamengo and Pyramids FC players had their data scanned before the match.
The core idea is similar to Referee View: better data and clearer communication will bridge the gap between technology decisions and audience acceptance.
One less-discussed but perhaps most important aspect of the partnership between FIFA and Lenovo is the development of an intelligent operations center. According to Romy Gai, this system connects real-time data from various departments, matches, stadiums, and broadcasters into a single operational control panel.
With a tournament taking place across three countries, involving over 180 broadcasters and approximately six billion viewers, operational coordination is crucial. This operations center is essentially the backbone of enterprise AI behind the AI technologies that FIFA has announced.
FIFA's removal of the role of local organizing committees is also a major change. This means FIFA must now directly take responsibility for many activities previously handled by national organizations. AI not only supports this decision, but it is also the very element that makes it feasible.
Football AI Pro is built on the Football Language Model, a specialized language model for football trained using data owned by FIFA. This is a very valuable asset. A conventional language model can answer questions about football, but a model trained with hundreds of millions of FIFA data points can produce accurate and in-depth analyses of the league that a general model would struggle to replicate.
The impact of this technology extends beyond the 2026 World Cup. FIFA has stated that in the future, Football AI Pro could be offered to fans as well, not just national teams. Furthermore, all 211 FIFA member federations are also planned to utilize this system.
If the model proves effective at the World Cup, it could become the foundation for a much larger project: bringing football data analytics capabilities to federations and leagues worldwide, especially those that currently lack sophisticated analytical tools.
That's the bigger story behind FIFA's announcements this week: the enterprise AI. The World Cup was just a test run. What FIFA builds upon that foundation is the real implementation process.