IBM on Tuesday issued a profit warning ahead of the start of the second-quarter earnings season: the company said revenues rose by only 1 percent in the most recent quarter, versus analysts' expectations of 5 percent. The market reacted sharply: IBM shares fell about 25 percent. CEO Arvind Krishna said the weak results reflect broader market dynamics rather than primarily company-specific failures.
What is shifting IT spending?
IBM reported that customers have shifted significant parts of their IT budgets in recent months toward investments that the AI boom has driven—servers, storage systems and memory chips. That shift has weighed on sales of IBM mainframes and software. IBM also cited cybersecurity concerns linked to Anthropic's Mythos model as a factor that may have disrupted corporate IT investment plans, with many clients prioritizing risk mitigation over other projects.
Investor fears and structural change in demand
IBM’s case illustrates how nervous investors are: a relatively modest profit warning triggered a large sell-off. The central market question is whether AI-driven infrastructure spending is already crowding out traditional IT expenditures, or whether that effect will materialize only over time. The Financial Times has noted that, until now, there was little clear evidence that AI was immediately displacing other IT investments; concerns such as the so-called “SaaSpocalypse” reflected longer-term risks to current software demand.
Analyst revisions and cost pressure
After the first quarter, Gartner raised its 2026 IT spending forecast in response to a sharp rise in data-centre spending. The prevailing expectation had been that AI would expand overall technology spending rather than do so at the expense of other IT investments. IBM’s warning suggests that narrative may be shifting.
On Wall Street, investors will closely watch two questions during the earnings season: how much the infrastructure investment wave has already crowded out traditional IT spending, and to what extent rising infrastructure costs will increase—and thereby constrain—demand for new AI services.
Timing and crowding-out risks
Krishna said customers are trying to buy AI hardware ahead of anticipated price rises, especially for memory chips. That behaviour creates two main risks:
- Timing uncertainty: advance purchases can make investment timing more unpredictable—companies that buy equipment early to avoid price rises may delay further spending while they absorb already-purchased hardware, which can amplify market volatility.
- Crowding-out effect: rising prices may divert larger shares of companies’ capital budgets to chip manufacturers rather than to creating additional compute capacity, which could constrain the supply side of AI.
Industry winners and losers
The IBM warning prompted further sell-offs in software stocks such as ServiceNow and Adobe. Consumer electronics makers may also be affected if rising component costs make PCs, game consoles and smartphones less attractive.
Chipmakers, by contrast, are enjoying strong margins: Nvidia’s gross margin is around 75 percent, and Micron reported a gross margin of 83 percent in its most recent quarter—more than double the level a year earlier. Those figures illustrate which vendors are benefiting from current demand for AI hardware.
Implications for the AI market
Performance gains from new AI accelerators and high-bandwidth memory partly offset price increases, and the per-token production cost for large language models could continue to fall. Nevertheless, fiercer competition for compute capacity is likely to keep AI service prices higher than they otherwise would be—precisely when demand is poised to accelerate.
Overall, IBM’s warning signals a potential structural shift in IT spending toward AI infrastructure, creating new risks for some incumbents and substantial opportunities for hardware suppliers.
