IBM Shares Suffer Worst Single-Day Drop in Over 25 Years Following AI Announcement
IBM shares experienced their most severe single-day decline in more than a quarter century on Monday, February 23, plummeting 13.2% to close at $223.35. This dramatic selloff erased approximately $40 billion from IBM's market capitalization and pushed the stock down more than 24% year-to-date. The catalyst for this market panic was an announcement from artificial intelligence startup Anthropic regarding its Claude Code tool's capabilities.
Anthropic's COBOL Modernization Announcement Triggers Market Panic
The stock collapse was directly triggered by a blog post from Anthropic claiming that its Claude Code artificial intelligence tool could handle the exploration and analysis work that makes COBOL modernization projects so expensive and time-consuming for enterprises. COBOL—Common Business-Oriented Language—is the decades-old programming language that forms the foundation of IBM's mainframe business.
Anthropic argued that AI tools can now compress what used to be multi-year, consultant-heavy migration projects into a matter of quarters rather than years. For IBM, which earns recurring revenue from mainframe hardware refresh cycles, software licenses, and COBOL-related services, this announcement represented a direct threat to a significant portion of the company's business model.
The Enduring Legacy of COBOL Programming Language
COBOL was created in 1959—the same year Alaska became a U.S. state—partly drawing on work by computing pioneer Grace Hopper. The language was specifically built for processing business data including payroll, transactions, and administrative records. Remarkably, sixty-six years later, COBOL continues to perform this function daily.
An estimated 95% of ATM transactions in the United States still rely on COBOL systems, while 80% of in-person credit card swipes are supported by the language. Hundreds of billions of lines of COBOL code remain in active production every day, powering critical systems at banks, airlines, and government agencies worldwide.
The COBOL Working Group of the Open Mainframe Project estimated in 2021 that approximately 250 billion lines of COBOL are still in use at businesses globally. Most of this code runs on IBM mainframes—the massive, customer-owned servers optimized for large-scale transaction processing.
The COBOL Developer Shortage and Modernization Challenges
The fundamental challenge with COBOL is not that it doesn't work—it performs extremely well for its intended purposes. The problem lies in the shrinking pool of developers who understand the language. Most contemporary computer science graduates are trained on Python, Java, and cloud-native architectures, viewing COBOL maintenance as career-limiting rather than career-building.
This has created an expensive talent bottleneck where organizations compete for a diminishing number of specialists while struggling to attract younger developers. During the COVID-19 pandemic, several U.S. states found themselves scrambling for COBOL programmers when unemployment systems—many still running on legacy code—buckled under sudden demand.
Banks have attempted multi-year migration projects to move away from COBOL, with some efforts resulting in widespread service disruptions and regulatory fines. The Internal Revenue Service only recently announced a transition from COBOL to Java. For most organizations, understanding legacy code has historically cost more than rewriting it—which explains why so much COBOL remains in production.
Anthropic's Detailed Claims About AI-Powered Modernization
In its Monday blog post, Anthropic framed the announcement as a direct solution to this modernization bottleneck. The company stated that Claude Code can map dependencies across thousands of lines of COBOL, document workflows, and flag risks that would take human analysts months to identify. Anthropic also released "The Code Modernization Playbook," outlining a phased approach where AI agents read through COBOL programs and JCL scripts, extract business logic, generate code translations to Java or Python, and create test suites—all within weeks rather than years.
Anthropic's broader argument is that legacy code modernization stalled because comprehension represented the real expense, not the rewriting itself. The company contends that artificial intelligence flips this equation by making analysis cheap and fast.
IBM's Mainframe Business Faces Direct Threat
IBM isn't merely a company that happens to use COBOL—it owns the mainframe platform on which the language predominantly runs. The company generates revenue from hardware, software licenses, and performance upgrades tied to COBOL workloads. IBM's modernization strategy has focused on connecting COBOL to modern technology—exposing COBOL programs as APIs, integrating them with cloud applications, and running them alongside Java and AI workloads—rather than eliminating the language entirely.
If an external AI tool can handle the heavy lifting of understanding, documenting, and migrating COBOL systems, it threatens a core component of what IBM sells. While COBOL won't disappear overnight, the consulting-heavy, multi-year modernization model that has sustained IBM and numerous large IT services firms—including companies like Infosys, TCS, and Wipro—could shrink significantly.
Broader Market Impact and Industry Disruption Fears
The market reaction extended beyond IBM alone. Indian IT benchmarks felt immediate ripple effects, with the Nifty IT index dropping nearly 4% on Tuesday, February 24, as fears of AI-driven disruption to legacy IT services spread through the sector.
IBM's Monday crash represents part of a much wider market rout affecting the software industry. On the previous Friday, cybersecurity heavyweights like CrowdStrike and Datadog slumped after Anthropic unveiled a separate security scanning feature in Claude Code. A major software exchange-traded fund has now shed 27% this year—its steepest quarterly decline since the 2008 financial crisis.
The pattern has become increasingly predictable: an AI company introduces a new capability, and investors rapidly dump legacy companies that capability threatens. The underlying fear driving this market behavior has been termed "vibe coding"—the idea that AI can now write functional software from plain-English prompts, leading investors to question the long-term pricing power of companies selling tools and services that developers might soon replace themselves.
Whether Anthropic's COBOL claims will hold up at enterprise scale—where decades of undocumented business logic, regulatory requirements, and organizational inertia complicate every migration—remains an open question. However, Wall Street isn't waiting to find out. The market has essentially decided that the era of expensive, multi-year legacy modernization projects is ending, with the only remaining debate being how quickly this transition will occur.
