COBOL Modernization Is a Talent Problem. Here Is What Hiring Leaders Need to Know.

Your mainframe team has a retirement wave coming. The engineers who built and maintained your core systems are leaving, and there are not enough qualified people ready to fill those seats. That is not a prediction. For most large U.S. organizations, it is already happening, and many talent acquisition leaders are only now recognizing the urgency.

COBOL is not going away. It runs more of the global financial infrastructure than most people realize, and modernizing it is now a strategic priority for banks, insurers, government agencies, and enterprises across the country. For technology and talent acquisition leaders, though, the most pressing challenge is not the technology. It is the people. Understanding who to hire, and how the candidate profile has changed, is the first and most consequential step.

This post breaks down what hiring leaders need to understand about COBOL modernization today, how AI has shifted the talent requirements, and how to build a team that can get the work done.

What Is COBOL Modernization, and Why Does It Matter Right Now?

At its core, COBOL modernization is the process of updating or migrating legacy COBOL-based systems to modern architectures, languages, or cloud platforms while preserving the business logic those systems carry. More broadly, COBOL software modernization encompasses everything from incremental code refactoring to full cloud migration of mission-critical applications.

That second part is what makes modernizing COBOL applications so difficult. These systems do not just contain code. In fact, they contain decades of accumulated business rules, regulatory logic, and institutional knowledge that was encoded by engineers who are no longer at the organization. Replacing that logic incorrectly can break processes that handle billions in daily transactions. In regulated industries, the stakes are even higher because errors can trigger compliance failures, not just operational ones.

Consider the scale. COBOL handles an estimated 95 percent of ATM transactions in the United States. According to IBM, more than 250 billion lines of COBOL code are in active production globally. Beyond that, industry estimates place 70 percent of all business transaction processing on COBOL infrastructure, representing roughly $3 trillion in commercial activity every day.

These systems work. The problem, however, is that the workforce needed to maintain and evolve them is shrinking fast. Without a deliberate hiring strategy, organizations risk losing the institutional knowledge that holds these systems together. For many enterprises, the modernization clock is already running.

The COBOL Modernization Talent Crisis Is Structural

The Numbers Behind the Shortage

The scale of the problem becomes clear when you look at the data. According to Deloitte and Forrester research cited by VentureBeat, 79 percent of companies identify acquiring mainframe talent as their top challenge. Beyond that, Data Center Knowledge reports that 71 percent of mainframe teams are currently understaffed, and 93 percent of organizations find qualified talent to be moderately to extremely challenging to source.

These figures reflect something structural, not cyclical. COBOL is taught at only a handful of universities today. New developers are not entering this space in meaningful numbers. Meanwhile, the generation that built and maintained these systems is retiring, and the institutional knowledge they carry leaves with them. No amount of compensation adjustment changes the fundamental math of a pipeline that is not being refilled. For organizations running large mainframe operations, that reality makes workforce planning for COBOL modernization a board-level concern, not just an HR one.

Why Traditional Hiring Cannot Keep Pace

Traditional mainframe staffing has relied on a narrow candidate pool. For many organizations, that pool has become so compressed that even sustained recruiting efforts cannot keep up with attrition. As a result, teams remain perpetually understaffed despite ongoing investment in hiring.

The deeper issue is that the conventional playbook, which involved recruiting experienced COBOL specialists and waiting 12 to 24 months for them to reach full productivity, no longer works at the pace modernization demands. Consequently, addressing that gap requires rethinking the talent profile entirely, not simply recruiting harder for the same one. The profile of the talent you need has shifted, and the good news is that AI is a significant reason why.

How AI Has Changed the COBOL Modernization Equation

From Manual Mapping to AI-Assisted COBOL Analysis

Historically, COBOL software modernization required enormous consulting teams working for years to manually map workflows, trace dependencies, and document systems with no existing documentation. Those timelines and costs kept many modernization projects stalled on the roadmap indefinitely, even when business leaders recognized the urgent need for change.

Fortunately, COBOL AI tools have changed that math. Platforms designed for COBOL AI programming, including Anthropic’s Claude Code, can now automate the exploration and analysis phases that once consumed most of the budget on these projects. In practice, teams using Claude Code for COBOL work can now approach modernizing COBOL applications in quarters rather than years, according to Anthropic. Beyond speed, COBOL modernization using AI significantly lowers the barrier to entry for engineers who lack deep legacy backgrounds, which directly expands your available talent pool and reduces your dependence on a narrow set of specialists.

Specifically, these tools can map dependencies across thousands of lines of code, surface undocumented workflows, identify migration risks, and generate documentation for systems that have gone undocumented for decades.

What AI Does Not Replace

Automation handles the analysis. However, human judgment handles everything else. The IEEE Computer Society emphasizes that software migration requires expert human oversight to validate outputs and confirm systems behave correctly in production. That principle applies with particular force in regulated industries, where a silent logic error can go undetected until it surfaces in a compliance audit or a production failure.

COBOL engineers understand regulatory requirements, business priorities, operational constraints, and risk tolerance in ways that AI tools cannot replicate. Human oversight is not optional on these projects. In fact, what COBOL and AI together accomplish is making it possible to staff that oversight with a broader range of talent than was previously viable. The AI handles the translation. Your team handles the judgment.