IBM shares slumped on Monday, February 23, 2026, after Anthropic said its AI could speed up COBOL modernization. The stock fell about 13%, marking IBM’s sharpest one-day drop since October 2000.
The sell-off weighed on the Dow Jones Industrial Average, which dropped more than 820 points in the session. Traders also tracked broader pressure on software-related stocks tied to AI disruption risk.
Investors focused on IBM’s exposure to COBOL workloads that run on IBM mainframes. Many regulated industries still rely on those systems for stable, high-volume processing.
IBM closed near $223.35, down about 13.15% for the day. That move ranked as IBM’s steepest daily fall in more than 25 years.
Market participants linked the move to fears that AI tools could lower the cost of modernizing legacy code. If customers modernize faster, they may rethink long-term mainframe demand and related services.
The risk-off tone extended beyond IBM during the same session. Cybersecurity stock names also fell as investors priced in faster AI-driven software development and testing cycles.
Anthropic published a blog post describing how “Claude Code” can support COBOL modernization work. It said teams can automate code exploration and analysis tasks that usually consume the most effort.
Anthropic also said AI can cut modernization timelines from years to quarters for some projects. It listed steps like mapping dependencies, documenting workflows, and identifying risks across large codebases.
The announcement landed during a period of heightened sensitivity across software stocks. Investors have watched several AI product releases reshape expectations for traditional software pricing power.
In Monday’s trading, that concern shifted toward legacy systems and modernization budgets. As a result, IBM stock price absorbed a large share of the day’s re-rating pressure.
Also Read: IBM to Acquire Confluent for $11 Billion to Strengthen Cloud and AI Strategy
COBOL remains common in core transaction systems, especially in banking and government operations. Many organizations kept COBOL because it supports predictable batch processing and high reliability.
Anthropic highlighted a staffing issue that complicates maintenance and upgrades. It said fewer schools teach COBOL, while many original system builders have retired.
That talent gap often forces companies to spend heavily on documentation and risk review before any migration. AI tools now target those early phases, which can shape budgets and timelines.
IBM has also invested in AI-driven approaches for modernization, including tools that help analyze legacy codebases. Even so, Monday’s sell-off showed how quickly markets can reprice disruption risk.