Artificial intelligence platforms are being compared to the legendary “Paul the Octopus” after accurately predicting key outcomes in the opening matches of the 2026 FIFA World Cup.
As the tournament kicks off across the United States, Canada and Mexico, several leading Chinese AI models, including Qwen, Doubao and DeepSeek, have collectively identified Spain as the favourite to lift the trophy, while tipping France captain Kylian Mbappé to finish as the tournament’s top scorer.
The comparison to Paul the Octopus comes after the famous German octopus gained worldwide fame for correctly predicting several major results during the 2010 FIFA World Cup in South Africa.
Now, AI-powered football predictions are emerging as a new trend among fans looking to forecast match outcomes.
Alibaba’s Qwen recently launched a dedicated football prediction assistant, allowing users to submit score predictions throughout all 104 matches of the expanded 48-team tournament.
Participants who correctly predict more than 80 matches and outperform Qwen’s own prediction accuracy stand a chance of winning cash prizes and AI-related products.
Qwen attracted widespread attention after successfully predicting Mexico’s 2-0 victory over South Africa in the opening match.
The AI system also forecast that South Africa’s aggressive style of play could result in a red card, based on factors including referee tendencies and match conditions.
Its prediction proved remarkably accurate when a South African player was sent off in the 49th minute, with the match eventually producing three red cards — a new record for a World Cup opening game.
Qwen also correctly predicted South Korea’s 2-1 victory over the Czech Republic in another early tournament fixture.
According to Qwen’s product manager Cheng Fei, the AI’s forecasts are generated using vast datasets that include team history, player performance, injuries and tactical trends.
The model also factors in environmental conditions such as altitude, geography, temperature, humidity, kick-off times and the typical playing conditions each team is accustomed to.
In the case of Mexico’s opening match, Cheng said the host nation’s advantage, combined with Mexico City’s altitude of more than 2,200 metres above sea level, played a significant role in the system’s prediction.
A review of several major Chinese AI platforms found that Qwen, Doubao and DeepSeek all ranked Spain as the leading contender for the World Cup title, assigning the team a championship probability ranging from 21 to 35 per cent.
France, England and Argentina were also consistently listed among the tournament favourites.
Despite the impressive forecasts, veteran football commentator Huang Jianxiang urged caution against placing too much faith in artificial intelligence.
He argued that football’s greatest appeal lies in its unpredictability and said that while AI may understand statistics, it does not necessarily understand the game itself.
The AI platforms also acknowledged their limitations, noting that factors such as tactical substitutions, individual mistakes, unexpected injuries and other unforeseen events can dramatically alter the outcome of a match.
“No AI can achieve 100 per cent accuracy,” the systems cautioned.
Sports consultant Zhang Qing said AI prediction campaigns are ultimately a form of digital marketing designed to capitalise on World Cup excitement, helping companies boost brand visibility, user engagement and data collection.
Technology analyst Wang Yan added that competition among AI companies has shifted beyond technical specifications and now centres on user attention, platform engagement, social media influence and long-term market adoption.
As the World Cup progresses, football fans around the world will soon find out whether AI can continue its impressive run — or whether the sport’s famous unpredictability will once again prove impossible to calculate.
