Skip to content
Artificial intelligence. Clearly. RSS feed
AI NEWS AI

News. Evidence. Perspective.

Build & CodeExplainer

How AI helps officials make an offside decision

Football’s offside technology combines player tracking, a sensor inside the match ball and the laws of the game. Each supplies a different part of the decision.

FIFA reported on 30 June 2026 that all 1,248 World Cup players had been scanned, giving its offside technology individual digital avatars for tracking and replays. AI helped create those avatars.[1]

AI combines camera tracking with ball-sensor information to locate players and the ball and generate automated offside alerts.[6]

The ball supplies a timing signal

The match-day Trionda ball contains a motion sensor in one of its four panels. According to adidas, it sends data 500 times per second and helps officials identify individual touches. Counterweights in the other panels balance the embedded electronics.[3]

That sampling rate means readings are separated by two milliseconds. It describes how frequently the sensor measures motion, not how quickly an entire offside decision reaches the referee. Ball information still has to be combined with player-position data and the rest of the officiating process.[3]

The process

The information an offside decision needs

The information an offside decision needs
QuestionRelevant evidence
When was the pass played?The connected ball helps identify the touch.Source 3
Where were the players?Tracking locates bodies within calibrated pitch geometry.Source 5
Which position matters?Law 11 defines eligible body parts and the offside position.Source 4
Was there an offence?Officials assess involvement, including interference without touching the ball.Source 4Source 2

A television picture has geometry to resolve

A camera projects a three-dimensional scene onto a flat picture. A foot on the grass and a shoulder above it cannot simply be treated as marks on the same ground plane. Establishing their positions requires a calibrated relationship between the camera view and the pitch.[5]

FIFA’s research from 2016 to 2019 identified pitch camber, camera-angle distortion, obscured players and body-part tracking as sources of error in virtual offside lines. A pitch can slope for drainage; one player can hide another. These are physical measurement problems that software must handle before the viewer sees a line.[5]

This helps explain why the clean broadcast graphic is the end of a measurement process. Its visual simplicity does not describe the work required to locate a moving body from camera images.[5]

The law distinguishes position from involvement

For a kicked pass, Law 11 uses the first contact with the ball to assess a teammate’s position. Relevant parts are the head, body and feet, excluding hands and arms. In the opponents’ half, the player must be nearer the goal line than both the ball and the second-last opponent to be in an offside position.[4]

Position alone is not an offence. The player must become involved, for example by playing the ball or interfering with an opponent. Blocking a goalkeeper’s view can matter without a touch. Players equally far forward can require different decisions if only one affects play.[4]

FIFA’s 2026 upgrade sends clear positional offside alerts directly to on-field officials. Its briefing also specifies a boundary: the technology does not determine interference by an offside player who never touches the ball. Better position and timing evidence can accelerate part of the process while that judgement remains with officials.[2]

Follow the evidence

Sources & further reading

  1. World Cup 2026: player avatars and technology update FIFA · 30 Jun 2026 · Accessed 12 Oct 2026
  2. World Cup 2026: offside and refereeing technology briefing FIFA · 3 Jun 2026 · Accessed 12 Oct 2026
  3. Law 11: Offside The International Football Association Board · Accessed 12 Oct 2026
  4. Offside Technology FIFA · Accessed 12 Oct 2026
  5. Club World Cup 2025: offside technology FIFA · 6 Jun 2025 · Accessed 12 Oct 2026