OpenAI publishes COBOL migration benchmark; safety gap remains
What happened
On October 26, The Verge reported on OpenAI's progress migrating COBOL code with AI. In tests, GPT-6.1 Sol kept behavior accurate 87% of the time at low reasoning effort and reached 98% with more reasoning effort. Experts said these benchmarks are still far from safely replacing critical legacy systems, and a gap remains between test scores and real system overhauls. The report also noted Anthropic has published a guide for code modernization with Claude Code but offers no migration performance benchmark.
Written by AI from the coverage · updated 48 minutes ago
Coverage
Follow the reports to see the story from different sides.
- The Verge · AICan AI help solve the COBOL code crisis?
OpenAI 报告了 AI 迁移 COBOL 代码的进展，受访专家指出这些评测距离安全替换关键遗留系统仍很远。其测试中，GPT-6.1 Sol 在低推理投入下的行为保持准确率为 87%，提高推理投入后达到 98%；Anthropic 也推出了使用 Claude Code 进行代码现代化的指南，但未提供迁移性能基准。
Heat over time
Not enough continuous observations to draw a trend yet.