AI search agents don't fail at searching, they fail at asking the right questions when queries get ambiguous

The new DiscoBench benchmark reveals that AI search agents struggle with ambiguous queries. Results show that these models perform worse, at 51.9%, when searching repeatedly without asking users for clarification. Even the best model only achieves 43% overall accuracy. Data indicates that removing ambiguity from queries can boost accuracy by up to 40 points. This suggests that enhancing models' ability to ask the right questions could significantly impact search outcomes. As reliance on intelligent search agents grows, this issue becomes increasingly critical.