About
AI Reasoning Benchmarks

AI Reasoning Benchmarks

1 entry in Legal Intelligence Tracker

1 Contributing Entry

New study shows OpenAI's GPT-5.5 failed to outperform o3 on law school exams

University of Maryland law professors have found that OpenAI's GPT-5.5 did not meaningfully outperform its predecessor, o3, on law school final exams—a finding that challenges assumptions about consistent improvement in newer AI models.

mail Subscribe to AI Reasoning Benchmarks email updates

Primary sources. No fluff. Straight to your inbox.

Also on LawSnap