The World Cup’s VAR Fights Are a Preview of Every Company’s AI Rollout Problem

The 2026 FIFA World Cup’s use of video assistant referees (VAR) has provided a fascinating case study in technology-assisted decision-making, echoing challenges faced by organizations rolling out AI systems across various sectors. As fans and players watched the drama unfold on the pitch, it was clear that VAR had not only failed to eliminate controversy but also created new ones.

The most striking aspect of VAR is its inability to live up to its promise of reducing subjectivity in refereeing decisions. Rather than eliminating disputes over judgment calls, VAR has shifted the focus from what the referee saw to how the process works. Similarly, AI adoption in organizations promises greater precision and accuracy but often raises expectations that are impossible to meet.

This phenomenon is not unique to VAR or AI. Many sports have adopted video review systems, with varying degrees of success. Baseball’s replay system, for example, has become a contentious issue in its own right, with disputes arising over the rules governing when technology can intervene. The same pattern emerges in organizations that adopt AI decision-making systems: while they promise to eliminate bias and increase efficiency, they often create new challenges related to accountability, transparency, and trust.

Subjectivity does not disappear with technology; it merely shifts. In VAR, arguments about refereeing decisions have moved from what the referee saw to how the process works. Similarly, AI systems create new forms of discretion, as organizations must decide which models to use, what data to train on, and what error rates are acceptable.

This shift in subjectivity has significant implications for trust in technology-assisted decision-making. Greater precision can reduce some types of uncertainty but often raises expectations that cannot be met. When technology fails to deliver on these promises, people become frustrated, and trust falls instead of rises. This finding is particularly relevant for AI adoption, where organizations frame the technology as a silver bullet for eliminating bias and increasing efficiency.

However, research suggests this framing is overly simplistic. The better question in AI adoption is not whether to automate or augment human judgment but which decisions should be made by each approach. Measurement problems are strong candidates for automation, as AI can process data faster and more consistently than people can. Interpretation problems require human-AI collaboration, where technology provides information, options, or recommendations, but people must think about the context of the decision.

The VAR experience offers a sobering reminder that technology cannot solve all our problems. While it may provide greater precision and accuracy in some areas, it also raises new challenges related to accountability, transparency, and trust. As organizations grapple with AI adoption, they would do well to learn from VAR’s failures and focus on creating systems that acknowledge the limits of technology while harnessing its potential to support human judgment.

The stakes are high: as we increasingly rely on AI decision-making systems, it is essential to get this right. The consequences of getting it wrong – decreased trust, increased controversy, and reduced effectiveness – will be far-reaching. By acknowledging the limitations of technology and working towards more nuanced solutions that balance precision with judgment, we can create systems that truly serve the public interest.

Technology must work in tandem with human judgment to produce fair and trustworthy outcomes. Anything less will only lead to further controversy and frustration – exactly what we’re trying to avoid with AI adoption in the first place.