AI models fail to profit from Premier League football betting
A study by startup General Reasoning revealed that artificial intelligence models from Google, OpenAI, and Anthropic incurred losses when placing bets on English Premier League football matches.
The experiment tested eight different AI systems that generated betting predictions based on historical data and team statistics, forecasting match outcomes and total goals. During testing, the AI models had no internet access and were given three attempts each to generate profits.
Each model started with £100,000 but ultimately, all AI agents ended the trial with financial losses. This indicates that even advanced AI systems are currently unable to reliably generate profit from sports betting without external information and additional training.
Startup General Reasoning focuses on AI research, emphasizing practical capabilities and limitations of contemporary AI systems in complex fields such as sports wagering.
The study demonstrates that despite technological advancements, AI still struggles to effectively predict sporting events for commercial gain. Greater success will require more sophisticated models or real-time data access.
Moving forward, as technology advances and AI incorporates more comprehensive datasets, its efficiency in forecasting may improve, but for now, sports betting remains a risky domain for AI systems.
