OpenAI has announced significant advancements in its AI models' ability to migrate legacy code, particularly COBOL, into modern programming languages. Their GPT-6.1 Sol model demonstrated an impressive 98% behavior preservation accuracy in tests, a substantial improvement over previous versions and competitors like GPT-6 Astra. This development could address the global crisis of aging, critical infrastructure code that is difficult and risky to update. Anthropic is also pursuing this market with its Claude Code agent, though without providing specific performance benchmarks. AI
IMPACT Could accelerate the modernization of critical legacy systems, reducing costs and risks associated with outdated codebases.
RANK_REASON AI model improvements for a specific software maintenance task.
- Anthropic
- Claude Code
- COBOL
- GPT 5.6 "Sol"
- GPT-6.1 Sol
- GPT-6 Astra
- GPT-6 Sol
- IBM
- Internal Revenue Service
- OpenAI
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →