Imagine a programming language created in the 1950s that still powers critical banking, healthcare, and retail systems worldwide. Now imagine artificial intelligence threatening to dismantle the multi-billion dollar industry built around maintaining these legacy systems. That’s exactly what happened this week when Anthropic’s announcement about its Claude Code tool for automating COBOL modernization sent IBM shares plunging over 13% in a single day – the company’s worst trading day in more than 25 years.
The COBOL Conundrum Meets AI
COBOL (Common Business Oriented Language) represents one of enterprise technology’s most persistent challenges. Developed in the 1950s, it still runs on thousands of mainframes handling trillions of dollars in daily transactions. These systems are notoriously difficult to modernize, often requiring “whole armies of consultants” working for years, as Anthropic noted in their blog post. The scarcity of COBOL developers and the immense “technical debt” these systems represent have created a lucrative maintenance industry dominated by IBM.
Anthropic’s breakthrough comes from Claude Code’s ability to automate the most time-consuming parts of COBOL modernization – the exploration and analysis work that previously took years. The company claims this could reduce modernization timelines from years to quarters. But here’s the twist: IBM itself introduced an AI assistant for COBOL-to-Java translation two and a half years ago, and CEO Arvind Krishna recently reported widespread adoption. So why the dramatic market reaction?
A Broader Market Anxiety
The IBM sell-off wasn’t an isolated incident. It occurred amid a broader market anxiety about AI’s disruptive potential across multiple sectors. According to Financial Times reports, US software stocks and private capital groups experienced significant selling pressure on Monday, with the S&P 500 falling 1.1% and the Nasdaq Composite losing 1.2%. Companies like Workday, CrowdStrike, and Datadog dropped more than 8%, while private capital giants including Ares, KKR, Apollo, and Blackstone fell over 6%.
UBS analyst Samantha Meadows explained the underlying concern: “Coding has become the first domain where AI demonstrably outperforms humans at scale and as a result, the software sector has emerged as the most immediate pressure point.” This sentiment was echoed across markets, with European wealth managers also facing pressure after US fintech Altruist unveiled an AI tax planning tool that can analyze complex financial scenarios in minutes.
The Geopolitical Dimension
Adding complexity to the AI disruption narrative are emerging geopolitical tensions. Anthropic has accused three Chinese AI companies – DeepSeek, Moonshot AI, and MiniMax – of conducting “distillation attacks” on its Claude models. According to TechCrunch and Financial Times reports, these companies allegedly used over 24,000 fraudulent accounts and generated more than 16 million exchanges to copy Claude’s capabilities in areas like agentic reasoning, tool use, and coding.
Dmitri Alperovitch, Chairman of the Silverado Policy Accelerator think-tank, commented: “It’s been clear for a while now that part of the reason for the rapid progress of Chinese AI models has been theft via distillation of US frontier models. Now we know this for a fact.” These allegations emerge amid ongoing US debates about AI chip export controls to China, with Anthropic arguing that such attacks require advanced chips and reinforce the need for export restrictions.
Beyond the Hype: Real Business Impact
While market reactions can be dramatic, the real question for businesses is practical: How will AI actually transform enterprise software? The Financial Times analysis suggests AI represents a fundamental architectural shift comparable to the move to cloud computing. Some executives publicly downplay threats while privately acknowledging AI’s disruptive potential.
ServiceNow CEO Bill McDermott recently mounted a spirited defense against “speculation that AI will eat software companies,” even as his company’s market value declined by $100 billion over the past year. Meanwhile, Salesforce reported $540 million in annual recurring revenue from AI (just 1.5% of total), and ServiceNow has $600 million in annual AI revenue – significant numbers but still small percentages of their overall businesses.
Rishi Jaluria, Chief Software Analyst at RBC Capital Markets, offers a sobering perspective: “SaaS companies have relied on price increases and cross-selling to maintain steady growth, but this will be harder to come by as IT budgets are diverted to AI.”
The Human Element in an AI World
Amid all the technological disruption, human factors remain crucial. In wealth management, executives argue that while AI can handle basic guidance, complex financial advice remains a “people business.” Mark FitzPatrick, CEO of St James’s Place, noted that advisers who don’t use AI “may well be replaced by those that do,” but emphasized that AI “was going to put the work that advisers do on steroids” rather than replace them entirely.
Similarly, in enterprise software, the transition may be more about augmentation than replacement. Aaron Levie, CEO of Box, pointed out that “the business model post-cloud was even better for the incumbents,” suggesting established companies with existing customer relationships might adapt successfully to AI integration.
Looking Ahead: More Than Just Code
The current market volatility reflects deeper questions about AI’s role in business transformation. Is this another technological shift that incumbents will navigate successfully, or does AI represent an existential threat to established software companies? The answer likely lies somewhere in between.
As Gina Mastantuono, CFO of ServiceNow, observed: “It takes time, and there’s some volatility along the way. But growth from AI is already clear and, with the transition moving much faster than previous tech shifts, evidence will continue to mount.”
What’s clear is that AI’s impact extends far beyond coding efficiency. From geopolitical tensions around intellectual property to fundamental questions about business models and human roles in automated systems, the enterprise AI revolution is just beginning. The companies that will thrive aren’t necessarily those with the most advanced AI, but those that can best integrate it with human expertise, existing systems, and evolving business needs.