Research Evaluates Quality and Readability of AI-Generated High-Intensity Interval Training Information

A new research report published on Bioengineer.org assesses the quality and readability of AI-generated information regarding high-intensity interval training. The study found that while the best-performing AI system produced good-quality, better-attributed answers, all systems generated content for readers at a level significantly above public-facing health communication targets. Perplexity had the highest median scores across various evaluation instruments, including DISCERN, EQIP, and the JAMA benchmark for transparency.