IBM stock price plunges 13%, wiping out $31 billion in market value, Claude's impact on COBOL business sparks AI replacement concerns

On February 25, IBM’s stock price plummeted by approximately 13% in a single day, wiping out nearly $31 billion in market value, prompting Wall Street to reevaluate its core business model. The immediate trigger for the sell-off was the technological breakthrough announced by Anthropic, whose Claude model is said to be capable of reading and modernizing traditional COBOL code—directly targeting IBM’s long-standing reliance on mainframe maintenance and enterprise consulting markets.

For years, IBM’s competitive moat in enterprise infrastructure has been built on deep services related to mainframe systems and COBOL language. Banks, insurance companies, and government databases still run大量 legacy code, making system maintenance, upgrades, and migrations long-term stable revenue sources. However, as AI-powered code migration tools become more mature, market concerns are growing that “AI-automated legacy system transformation” could shorten high-cost consulting project cycles and reduce dependence on traditional service providers.

From a market structure perspective, companies are accelerating cost reduction and efficiency improvement strategies, with demand for automated software reengineering rising sharply. If Claude can reliably handle complex legacy code and generate modern architecture solutions, it could significantly lower the barriers to COBOL system migration. Investors have quickly priced in the risk of “AI disruption to enterprise IT services,” leading to a concentrated wave of selling.

It is noteworthy that billions of lines of COBOL code still run in the global financial system, involving payroll systems, insurance platforms, and critical government infrastructure. Historically, due to technical complexity, high compliance requirements, and migration risks, enterprises preferred long-term outsourcing of maintenance services, a trend that has sustained IBM’s consulting and infrastructure profit margins. Now, if AI-assisted code reengineering can be scaled effectively, companies might shift toward faster, lower-cost modernization paths.

However, industry experts point out that key system migrations still require high reliability verification and security audits. The accuracy and compliance of AI tools when handling extremely large codebases remain core challenges. Therefore, in the short term, a hybrid model of “AI + traditional services” is more likely than complete replacement. For IBM, whether it can establish technological leadership in enterprise AI modernization solutions will be a critical factor influencing its stock performance and competitive positioning in enterprise technology.