A study evaluated several AI systems on ten research-level mathematics problems created by prominent mathematicians. The results include AI-generated solutions, human solutions, and referee reports, offering a detailed assessment of AI performance in solving advanced mathematical problems.