Humans Outclass AI During Yearly Math Olympiad
Human participants came out on top against advanced AI systems developed by OpenAI and Google at this year’s International Mathematical Olympiad (IMO). However, for the first time, the AI models reached gold-medal scores—a sign of how rapidly they’re catching up. The IMO is an elite math contest for students under 20. While five students managed to score a perfect 42 out of 42 points, neither of the AI systems reached that level. Around ten percent of human contenders got gold-level scores. Google revealed that its updated Gemini chatbot solved five of the six problems presented at the 2025 IMO, held…