IBM Share Price Plunges 13% Amid AI Threat to Mainframe Legacy Business

The IBM share price plunged 13.2% on 23 February 2026, closing at $223.35. The sharp sell-off marked the company’s steepest single-day decline since 18 October 2000. The trigger came shortly after AI startup Anthropic announced that its Claude Code platform can modernise COBOL, the legacy programming language that forms a critical backbone of IBM’s mainframe ecosystem.

For investors monitoring IBM shares, the development signals a potentially significant shift. Artificial intelligence may now be capable of disrupting one of IBM’s most dependable and recurring revenue streams.

What Triggered the IBM Share Price Drop?

The decline in the IBM stock was sparked by a blog post from Anthropic. The company claimed that its AI system, Claude Code, can analyse and modernise COBOL code at scale.

COBOL, introduced in 1959, remains widely used across:

• Global banking systems
• Airlines and logistics networks
• Insurance companies
• Government departments

According to industry estimates from research groups such as Gartner and Micro Focus, there are still hundreds of billions of lines of COBOL code running mission-critical systems worldwide. Many of these systems operate on IBM mainframes.

For decades, maintaining and modernising COBOL applications required specialised consultants. This created predictable, recurring revenue for IBM through long-term service contracts and infrastructure support.

If AI can significantly reduce modernisation costs and time, that revenue model faces pressure. Investors reacted swiftly, sending IBM stock price sharply lower.

Why COBOL Matters to IBM’s Business Model

Mainframe Revenue Stability

IBM’s Z series mainframes continue to serve industries where uptime, security, and transaction reliability are non negotiable. While the company has diversified into hybrid cloud and enterprise AI, mainframes still generate billions of dollars through hardware sales, licensing, and support contracts.

The concern reflected in the falling IBM share price is not that mainframes will vanish overnight. Instead, investors worry that AI driven automation may weaken the traditional lock in created by legacy system complexity.

Historically:

• COBOL modernization projects were slow
• Skilled COBOL developers were scarce
• Migration risk was high

These barriers protected IBM’s ecosystem. AI tools may lower those barriers.

IBM Shares Performance in 2026

After the 13.2% decline, IBM stock are now down roughly 25% year to date. Investors are reassessing several key factors:

• The durability of IBM’s consulting revenue
• The long term relevance of legacy modernization contracts
• The speed at which AI can automate enterprise IT

The reaction mirrors broader concerns in the software and services sector, where generative AI tools are compressing service margins and shortening project timelines.

Is the Market Overreacting?

Although the drop in IBM market value appears dramatic, a balanced perspective is important.

Enterprise Systems Do Not Change Overnight

Large banks, insurers, and government agencies rarely overhaul mission critical systems quickly. Even if AI can analyse COBOL code efficiently, migration still requires regulatory compliance checks, security audits, and operational testing. These transitions often span years rather than months.

IBM’s Own AI Capabilities

IBM is not standing still in artificial intelligence. Through its enterprise AI platforms, including Watson and its broader AI portfolio, the company has invested heavily in automation and machine learning. IBM is also embedding AI capabilities directly into its mainframe offerings.

This means AI disruption could also open new revenue streams for IBM in advisory, integration, and AI driven infrastructure optimisation.

Hybrid Cloud Strength

IBM’s hybrid cloud strategy, strengthened following its Red Hat acquisition in 2019, provides exposure beyond traditional mainframe services. Modern enterprise architectures increasingly integrate mainframes with cloud native systems rather than replacing them entirely.