OpenAI says its new models show major performance leaps in low-resource legacy languages like COBOL, Fortran, and Visual Basic, used by banks and governments

AI can now help modernize the critical but outdated code banks and governments rely on.

OpenAI has announced advancements in its AI models that significantly improve their ability to understand and process legacy programming languages, including COBOL, Fortran, and Visual Basic. These older languages remain critical for many core systems in sectors like banking and government. The improvements focus on enabling developers to better maintain and modernize these existing systems, a process that involves deciphering decades of business logic. This development aims to bridge the gap between legacy infrastructure and modern software development practices, facilitating system upgrades and ensuring the continued operation of essential services that rely on these foundational languages. The company highlighted that progress in this area helps preserve the functionality of existing systems while allowing for modernization.

This capability upgrade signals a growing trend of AI tackling increasingly complex and specialized software development challenges, moving beyond modern languages to address the vast installed base of legacy code. Companies with significant investments in systems built on COBOL, Fortran, or Visual Basic can now look to AI tools for more effective migration and modernization strategies. This could reduce the cost and risk associated with these often-daunting projects, potentially accelerating digital transformation for a large segment of the global economy that has been constrained by these older technologies.

For developers and IT departments, this means a potential new avenue for addressing technical debt and extending the lifespan of critical applications. It affects organizations that may have struggled to find developers skilled in these niche languages or faced high costs for maintenance and updates. The ability of AI to parse and understand the intricacies of decades-old business logic could unlock significant efficiency gains and enable new features or integrations that were previously impractical or too expensive to implement.