Researchers tested OpenAI’s Codex on real-world data analysis tasks spanning business and scientific domains. The findings show Codex performs well on tasks with clear rules, visible data, and explicit schemas, achieving 83.7% accuracy on medical calculations and reliably completing file reading and statistical analysis. However, the model struggles with implicit metric definitions, multi-table joins, and aligning business concepts with database entities, leading the authors to recommend treating Codex as a collaborative assistant rather than an autonomous analyst.
