A New Benchmark in AI Capability
Anthropic's Claude 5, released this week, has achieved a 99.8% score on the Uniform Bar Examination, surpassing the average human performance by a significant margin. The model also scored in the 99th percentile on the USMLE (medical licensing exam) and CFA Level III.
These results represent a quantum leap from previous models and have reignited debates about the role of AI in professional services.