Can AI help solve the COBOL code crisis?

What happened

OpenAI announced a new initiative to tackle the ongoing COBOL code crisis plaguing essential systems in government and industry. COBOL, a programming language created in 1973, still runs critical operations at the IRS, banks, airlines, and many government infrastructures worldwide. These legacy systems rely on aging code that few developers understand or dare to modify for fear of causing system failures. OpenAI aims to use AI to interpret, update, and possibly migrate COBOL code to more modern languages and platforms.

Why it matters

This crisis creates a silent but pressing risk for organizations that depend on COBOL systems to process trillions in transactions and sensitive data daily. Skilled COBOL programmers are retiring or leaving the workforce, and rewriting millions of lines of code manually is costly and error-prone. If AI can accelerate understanding and migration of COBOL code, it could reduce system downtime, lower maintenance costs, and prevent catastrophic outages caused by fragile legacy infrastructure. Simply put, AI’s ability to automate legacy code transformation can unlock massive operational resilience and cost savings.

What to watch next

Tracking OpenAI’s progress on practical COBOL code migration tools will be key. The industry will gauge how effectively AI models handle complex legacy logic and produce reliable, maintainable output. Success will pressure traditional tech vendors and legacy software integrators to adopt AI solutions or risk obsolescence. Regulators and infrastructure operators may also start demanding AI-assisted validation to manage risks from unknown legacy code. The coming months will show if AI moves from promise to practical enabler in this overlooked but critical technology bottleneck.

AI Quick Briefs Editorial Desk