IBM under AI pressure

Shares of IBM lost about 13% of their value on Monday — one of the most notable single-day drops in recent times. The formal trigger was the rapid spread of the Claude Code tool from Anthropic, which, according to developers, can automate the modernization of legacy software written in COBOL. For the market, this sounded like a warning signal.

To understand the scale of the reaction, it is important to remember that COBOL is not a museum artifact but still a living infrastructure of the global financial system. By various estimates, about 95% of ATM transactions worldwide rely on systems built in this language. Banks, insurance companies, and government institutions have built critical processes on it for decades. The problem is that COBOL specialists are becoming fewer, and modernization of such systems traditionally requires enormous costs, multi-year projects, and large consulting teams.

This is where the AI factor appears. If tools like Claude Code can truly analyze old code, rewrite it into modern languages, or adapt it to new architectures within minutes, the traditional business model of corporate modernization may change. What used to cost millions of dollars and take years could theoretically become cheaper and faster. For clients, this sounds like an efficiency revolution. For traditional IT service providers, it looks like a threat to margins.

Investors primarily viewed this as a risk to IBM’s services business. The company has earned revenue for decades from maintenance, support, and transformation of large enterprise systems. If AI tools significantly reduce manual work and the need for large consulting teams, part of the revenue may come under pressure. Markets dislike uncertainty, and when it comes to fundamental changes in demand structure, reactions are usually sharp.

The IBM situation is only part of a broader trend. Generative AI is already affecting the entire software, consulting, and cybersecurity sectors. Companies whose revenue depends on man-hours and complex integration projects face questions: if algorithms can perform a large portion of tasks faster and cheaper, business valuations may be revised. This does not necessarily mean an immediate drop in revenues, but it changes expectations regarding future growth rates and profitability.

At the same time, it is important to note that the stock decline was not caused by weak financial results or deteriorating operating performance. It was a reaction to fear — the fear that the technological wave of AI could redistribute profits within the industry. In such situations, markets often price in the worst-case scenario long before it becomes reality.

However, IBM also has its own artificial intelligence strategy. The company is actively promoting the Watsonx platform, focused on enterprise AI solutions, data management, and process automation. Instead of ignoring the technological shift, IBM is trying to integrate into it. Moreover, cooperation with Anthropic could allow the company not only to use new tools but also to participate in shaping standards for their use in the corporate environment.

For large clients, not only speed and cost of modernization matter, but also security, data control, and regulatory compliance. In this segment, IBM still has strong positions thanks to long-term contracts and deep integration into client infrastructure. Even if AI accelerates the technical part of work, strategic consulting, risk management, and enterprise-level implementation remain complex tasks.

Thus, the current stock decline looks more like a reflection of changing expectations rather than evidence of business weakness. Investors should closely monitor how IBM will monetize its AI solutions, whether it can integrate new tools into existing service offerings, and retain clients running on legacy systems. The question now is not whether demand for modernization will disappear, but who will earn from this modernization in the AI era.