From Paul the Octopus to AI: Chatbots predicting FIFA winners
What's the story
As the excitement builds for the 2026 FIFA World Cup, fans are turning to artificial intelligence (AI) chatbots to predict potential winners. The trend mirrors that of Paul the Octopus, who gained fame in 2010 for predicting winning teams by eating from containers marked with flags. This year's tournament is special as it's the first one with widespread access to generative AI chatbots like OpenAI's ChatGPT and Anthropic's Claude.
Predictions
Chatbots split on World Cup winner
Bank of America analysts found their Microsoft Copilot chatbot equally favored Spain and France, with France predicted by around 40% of fans. Tech news site Tom's Guide also asked Gemini, ChatGPT, and Perplexity for predictions. All three picked Spain as the likely winner with France as the second choice. However, Chinese chatbots DeepSeek and Qwen tipped Argentina instead.
AI impact
Research underway to track AI predictions
The growing interest in generative AI systems' football foresight is evident from institutions ranging from banks to universities testing these technologies. Researchers at Germany's Ludwig Maximilian University (LMU) are conducting a more scientific analysis, tracking the accuracy of each AI model's game-by-game predictions on a public website called LLM SoccerArena. "The question of whether language models can reliably support real decision-making situations is critical," LMU management researcher Stefan Feuerriegel said.
Information integration
Adapting to dynamic situations
The Munich researchers are testing the AI's forecasts based on their internal knowledge and its ability to integrate online information about injuries, squad selections, and even betting markets into their predictions. This is a key part of the study as it shows how well these systems can adapt to dynamic situations and uncertainty in real-world scenarios.