AI wipes $31 billion off IBM over COBOL coding

In a single trading day, IBM lost market value comparable to the capitalization of several major technology companies. The giant's stock plummeted, erasing 31 billion dollars from stock market accounts — the worst result in nearly three decades. The catalyst for the panic was, seemingly, an archaic programming language called COBOL and the quiet yet devastating ability of modern AI models to work with it as confidently as an experienced programmer with thirty years of experience.

To understand the scale of what is happening, context is needed. COBOL — a language created back in 1959 — remains the backbone of the world's financial infrastructure. According to various estimates, systems written in it process more than three trillion dollars in financial transactions daily: banks, insurance companies, government agencies, pension funds.

The problem is that COBOL specialists are catastrophically scarce — their average age has long exceeded sixty, and universities stopped training such programmers back in the last century. It is precisely in this gap between outdated technology and talent shortage that IBM built its business for many years: expensive consulting, long-term contracts for maintaining legacy systems, manual code migration — all of this brought stable and quite impressive income.

Now this model is crumbling. Generative AI models have learned to understand COBOL with an accuracy that industry specialists recently considered unattainable. Modern systems are capable not only of reading and explaining COBOL code, but also of translating it into modern programming languages — Java, Python, Go — while preserving business logic that is sometimes not documented anywhere and exists only in the code itself. What used to take years of work by hundreds of engineers and hundreds of millions of dollars in budget can now be accomplished fundamentally differently: faster, cheaper, and without dependence on the exclusive competencies of a single vendor. Investors understood this immediately.

The market's reaction exposed a deep contradiction in IBM's position. The company itself actively invested in AI and publicly declared technological transformation, yet investors suddenly realized that this transformation is hitting IBM first and foremost. If clients can solve legacy system migration and modernization tasks themselves — or with the help of cheaper AI tools — why would they pay IBM millions for the same thing? The company's stock plummeted, recording the worst single-day drop since 1998. This is not panic from an irrational crowd — this is a sober recalculation of the company's future cash flows, whose business model rested on technological complexity that AI is now beginning to dissolve.

The consequences extend far beyond one company. The IBM story is a vivid demonstration of how AI destroys not only low-skilled labor, but also highly qualified niche expertise, for which enormous sums were traditionally paid precisely because they were rare and difficult to reproduce. The corporate IT consulting market — one of the most stable and high-margin segments of the technology industry — faced for the first time a direct and measurable threat from generative models. Analysts have already begun to revise valuations not only of IBM, but also of other major consulting structures with similar dependence on the complexity of outdated systems.

It is also significant that the blow came precisely from the direction of COBOL — a technology that for decades was called "too old for automation" and "too specific for AI". This debunks one of the main arguments in favor of the resilience of traditional IT consulting: supposedly, corporate legacy systems are so unique and convoluted that they don't lend themselves to any automation. It turns out, they do. And this changes not only IBM's position, but the entire logic of pricing in the corporate IT transformation market.

IBM faces the task of rethinking exactly what it is selling to clients in a world where the barrier to entry for COBOL expertise has dropped sharply. A company with a century and a half of history has more than once managed to reinvent itself — from tabulating machines to mainframes, from hardware to services, from services to cloud. The question is whether it will manage to do so quickly enough, before the market has recalculated its value completely.