Ted Hisokawa
Oct 14, 2025 17:06
GitHub Copilot and AI brokers are revolutionizing the modernization of legacy programs, making out of date COBOL programs accessible to right this moment’s builders, in accordance with GitHub’s insights.
In an period the place legacy programs nonetheless underpin important world infrastructures, GitHub Copilot and AI brokers are stepping as much as modernize these outdated frameworks. In response to a report by GitHub, these applied sciences are notably efficient in making COBOL programs, which have been in use for over six a long time, accessible to up to date builders.
The Problem of COBOL
COBOL, a programming language that predates the web, stays integral to many banking, insurance coverage, and authorities programs. Regardless of its significance, discovering builders expert in COBOL is more and more troublesome as the unique consultants retire. This has left a good portion of the world’s 200 billion strains of COBOL code in want of modernization.
AI-Powered Options
Julia Kordick, a Microsoft World Black Belt, has been on the forefront of utilizing AI to modernize COBOL programs without having to be taught the language. By leveraging AI for reverse engineering, her workforce can extract and doc enterprise logic, determine dependencies, and supply a complete understanding of legacy codebases.
A Systematic Method to Modernization
Kordick’s workforce employs a three-step framework for modernization: code preparation, enrichment, and automation aids. GitHub Copilot assists in reverse engineering code to grasp its performance, whereas AI brokers enrich this understanding by translating and analyzing the code’s construction.
Automation aids additional scale this course of, permitting for the era of name chain maps and test-driven modernization plans. These steps not solely streamline the modernization course of but in addition make sure that human consultants stay integral to validation and oversight.
Actual-World Utility and Limitations
Whereas AI supplies vital efficiencies, Kordick warns that it’s not a panacea. Human experience is crucial for validating AI-generated evaluation, and the distinctive complexities of every COBOL codebase imply that full automation shouldn’t be but possible. However, AI has already reworked legacy modernization from a frightening activity right into a manageable mission.
Open-Supply Framework for COBOL Modernization
GitHub and Microsoft have open-sourced a framework constructed with Microsoft Semantic Kernel, providing a sensible place to begin for organizations seeking to modernize their legacy programs. This framework consists of a number of specialised brokers that deal with varied facets of the modernization course of, from dependency mapping to value monitoring.
The framework is designed to maintain organizations accountable for their mental property whereas permitting inner groups to be taught and adapt to new enterprise logic as modernization progresses.
For extra detailed insights, go to the total article on [GitHub](https://github.weblog/ai-and-ml/github-copilot/how-github-copilot-and-ai-agents-are-saving-legacy-systems/).
Picture supply: Shutterstock
