On the road to the World Cup

The World Cup is about to kick off: this time, AI is taking to the field too

The first major sporting event of the artificial intelligence era could redefine the concept and use of big data

by Marco Barlassina

10 June 2026

4' min read

Translated by AI

Versione italiana

Key points

Even bigger big data
The Exabyte Era

4' min read

Translated by AI

Versione italiana

Lionel Messi walks steadily towards the penalty spot. He stops. A sigh, a short run-up, and the ball finds the back of the net.

It is 4.23 pm Italian time on 18 December 2022, a date that will go down in history as “the day that shook the internet”. That moment in the final of the 2022 FIFA World Cup in Qatar between Argentina and France was in fact watched live by around 1.5 billion people across all available platforms worldwide: almost 20% of the world’s population gathered in a single moment in front of televisions, the internet, streaming services, smartphones and other devices. It is estimated that that single minute generated over 500 terabytes of streaming video traffic and around 800 million interactions on social media. In practice, those moments alone would have accounted for between 4% and 5% of global internet traffic.

With just a few days to go before the kick-off of the 2026 World Cup in North America (kicking off on Thursday 11 June at 9.00 pm with the match between Mexico and South Africa), we can be reasonably certain that that record will be broken. Firstly, because four years ago most people still followed footballvia traditional television, whereas today the focus, in terms of how the event is consumed, is shifting towards social media and digital platforms.

According to FIFA, the World Cup in Qatar generated a total of 5 billion media interactions. The audience figures for traditional television and digital streaming were broadly equivalent (2.9 billion versus 2.7 billion), whilst social media was not far behind with 2.2 billion. Bank of America, which has dedicated a massive report produced by its analysts to the upcoming World Cup, believes that this year’s final could account for up to 7% of global internet traffic, but above all points out that the figures seen so far were recorded before the surge of artificial intelligence.

Big data gets even bigger

“The 2026 World Cup will be the first tournament in which data itself becomes a primary product,” write the analysts. “All teams will have access to real-time artificial intelligence models capable of analysing millions of data points, as well as three-dimensional simulations of events to build their strategies.” And whilst in the past the wealthiest teams enjoyed a competitive advantage, in 2026 AI will democratise access to data. “All national teams are already using artificial intelligence systems to process performance metrics in real time that didn’t even exist during the previous tournament. We won’t simply be watching a football match: we’ll be witnessing a massive simulation in real time, in which the physical world will be constantly transformed into data at a rate of several petabytes a week’.”

According to SanDisk’s estimates, the 2026 World Cup will be the most AI-driven and data-intensive ever organised, exceeding the threshold of 90 petabytes. This is 45 times the amount of data generated during the 2022 World Cup in Qatar. To put this into perspective, 1 petabyte is equivalent to approximately 13.3 years of high-definition video. This means that the event will produce the digital equivalent of over 1,200 years of HD content in just over a month.

IL DIZIONARIO DEI BIG DATA

Un byte è l’unità di misura fondamentale dei computer. Quando si calcolano i suoi multipli, l’esponente indica il numero di zeri da aggiungere. Ad esempio, un Megabyte è 10^6, il che equivale a un 1 seguito da 6 zeri: ossia un milione di byte

The era of exabytes

However, the estimate of 90 petabytes excludes social media, selfies, interactions via mobile devices and other platforms. Taking these sources and AI-generated data into account, Bank of America estimates that ‘total data creation could reach a record level of 2 exabytes, equivalent to around 45,000 years of 4K video’.

But where will all this data come from? Primarily from Football AI Pro, a tool through which FIFA and Lenovo have integrated generative AI into the very heart of the game. It is an advanced assistant for match analysis. According to FIFA’s website, it is ‘the first tool of its kind designed for analysts and technical staff, focusing on in-depth analysis and report generation based on official match data. It combines AI agents capable of querying structured match data (event and tracking data) with the FIFA Football Language model, video and other data sources to provide tactical analysis, performance analysis and strategic recommendations in a fast, reliable and structured manner'. The system will analyse hundreds of millions of FIFA data points and over 2,000 football metrics, including pressing, movements, tactics and transitions. Each team will therefore have its own AI model that will analyse all available information in real time.

We will then see the return to the limelight of the metaverse, in the form of three-dimensional avatars. BofA explains further: ‘Each player will be digitally scanned in about a second to create a high-precision 3D avatar used in decisions regarding offside and VAR. Not to mention that each of the 16 host stadiums will have a real-time digital twin used to monitor crowd flows, security and even biometric data from players’ wearable devices.”

A purely quantitative figure, then, goes a long way towards explaining the increase in data generation: with 48 teams and 104 matches, compared to 32 teams and 64 matches in 2022, the volume of streaming traffic is bound to be unprecedented.