AI disruption rippled through global markets this week after Anthropic — the AI research company behind Claude — revealed capabilities in its Claude Code tool to understand, analyse and modernise COBOL, a decades‑old programming language deeply embedded in enterprise systems.

COBOL (Common Business‑Oriented Language) runs critical infrastructure across sectors like banking, insurance and government, and continues to power legacy systems on IBM mainframes. Traditionally, updating COBOL systems has been a labour‑intensive, consultant‑heavy process that requires significant expertise and time.

Anthropic’s blog post suggested that Claude Code can dramatically accelerate this work by automating code exploration, dependency mapping and documentation — tasks that previously took years and large teams of engineers.

The market reaction was swift. On February 23, 2026, IBM’s share price dropped roughly 13.2 % — its worst single‑day performance since October 2000 — wiping out an estimated tens of billions of dollars in market value as investors reassessed the impact of AI tools on legacy technology services.

Concerns aren’t limited to IBM. Broader software and cybersecurity stocks also weakened as markets priced in the possibility that AI might upend established tech revenue models by reducing reliance on traditional consulting, modernization and IT services.

While IBM continues to develop its own AI offerings and remains a significant player in enterprise computing, Thursday’s sell‑off underscored how fast‑evolving AI capabilities — especially in automating complex technical tasks — are reshaping investor expectations and industry dynamics across the tech sector.