IBM’s storied stability cracked wide open as shares nosedived 13.2% to $223.35, the steepest daily fall in 25 years, fueled by AI’s encroachment on the colossal legacy software sector.
Investors, watching a 25% year-to-date erosion, are betting big on disruption. Anthropic’s announcement that its Claude Code AI masters COBOL—decoding, modernizing, and securing systems built in the Eisenhower era—ignited the sell-off.
This 70-year-old language powers irreplaceable infrastructure: 95% of U.S. ATM transactions, airline reservations, insurance claims, and government databases. IBM has thrived on the maintenance quagmire—slow upgrades demanding rare COBOL wizards and hefty fees.
But Anthropic envisions AI transforming this $1 trillion industry. It can dissect billions of code lines, reveal hidden dependencies, auto-generate docs for forgotten systems, and preempt risks that once took months to uncover. Modernization costs, historically prohibitive, plummet with AI efficiency.
As expertise evaporates, AI steps in where humans falter. IBM’s mainframe dominance, tied to COBOL servicing, suddenly looks vulnerable. The blog post crystallized fears: AI doesn’t just assist—it supplants the expensive, error-prone human processes that sustained IBM’s enterprise fortress.
This plunge signals broader tremors. Enterprises worldwide may accelerate migrations, starving legacy vendors of recurring income and heralding AI’s dominance in software rebirth.