Can AI solve the COBOL code crisis?
openai
| Source: The Verge | Original article
OpenAI proposes using AI to tackle the global COBOL code crisis that endangers critical systems in finance, government and transportation.
OpenAI announced on October 19 that it is exploring the use of generative‑AI to address the growing “COBOL crisis” – the shortage of engineers able to maintain the decades‑old code that still underpins tax agencies, banks, airlines and critical public‑sector systems worldwide. The company said its models can analyse, document and suggest changes to legacy COBOL programs, potentially lowering the risk of outages caused by the dwindling pool of specialised developers.
The issue is acute: many essential services rely on software written in the 1960s language, and organisations struggle to find staff who understand its quirks. Manual rewrites are costly and error‑prone, while the threat of a single bug cascading through financial or governmental infrastructure is a constant concern. If AI can reliably surface hidden dependencies, generate test cases or propose refactorings, it could extend the useful life of existing systems and free scarce human talent for higher‑value work.
OpenAI’s move follows a wave of AI‑driven code tools that have begun to assist developers with modern languages, from bug‑detection platforms to context‑engineering frameworks. The next steps will reveal whether the technology can meet the stringent reliability and security standards demanded by legacy environments. Watch for pilot deployments with public‑sector partners, regulatory feedback on AI‑generated code in mission‑critical contexts, and any open‑source or commercial extensions that bring the capability to a broader set of organisations.
Sources
Back to AIPULSEN