What if up to 80 cents of every dollar in your technology budget is used merely to keep aging systems on life support rather than driving enterprise growth? For executive leaders across North American banking, healthcare, and retail enterprises, legacy software has evolved from a quiet technical debt item into a a fundamental strategic paralysis.
A 67-year-old programming language called COBOL still quietly powers the financial world. It processes over $3 trillion in banking transactions every day, supports an estimated 95% of ATM transactions and remains embedded in more than 40% of U.S. banking systems.
The original engineers who built these systems are retiring, while banks are paying a premium to keep the remaining COBOL experts on board and training younger developers in a language the industry has spent decades trying to replace.
But replacing COBOL is far from simple. Commonwealth Bank of Australia spent five years and roughly $750 million on a core banking modernization effort. Across the industry, similar migrations have taken years, cost hundreds of millions, and sometimes failed to deliver the expected results.
Why? Because COBOL is rarely the real problem. Decades of business logic, data structures, hardware dependencies, and interconnected systems are.
Now, AI promises a shortcut: translate millions of lines of COBOL into Java or Python in days.
In February 2026, Anthropic announced that Claude Code could analyze and translate legacy COBOL into modern languages. IBM’s stock plunged 13.2% in a single day, wiping roughly $40 billion from its market value as investors feared AI could disrupt its mainframe business.
But this reaction revealed a bigger misconception: Translation is not modernization.
Changing the programming language does not automatically change the architecture. The real opportunity for AI is not simply rewriting legacy code, but helping enterprises understand what should be preserved, redesigned or replaced before modernization begins.
Discover how much technical debt is capping your operating margins. Make a self-assessment to evaluate your portfolio’s readiness and establish a clear modernization roadmap.
AI nuclear energy background, future innovation of disruptive technology
Many corporate technology budgets conceal significant financial friction within standard operational line items. Benchmark data indicates that large enterprises spend up to 80% of total IT budgets merely maintaining legacy applications, leaving barely 20% for innovation or growth. In sectors like banking and healthcare, legacy maintenance frequently absorbs 70% of total technology expenditure.
Four primary operational drivers compound these hidden costs:
Transitioning to an AI augmented modernization model restructures enterprise cost dynamics. Modernization shifts from a high-risk rewrite into a controlled, value driven evolution.
Operating on rigid legacy systems directly threatens corporate valuation, market agility, and long-term scaling targets. While core modernization is no longer negotiable, the path forward requires more than just deploying technology; it demands implementing AI the right way—leveraging smart architectures like sidecar modernization to preserve embedded business logic without operational disruptions.
Navigating this architectural evolution requires a balanced approach of automated intelligence and deep engineering expertise. As your strategic partner, CMC Global helps you manage this transition seamlessly, turning legacy constraints into clean, scalable growth.
Evaluate your application readiness with our experts to map your 90-day modernization roadmap.