IBM has warned that enterprise customers are redirecting technology budgets toward AI infrastructure, a shift that disrupted its second-quarter results and triggered a selloff across software stocks.

The announcement sent IBM shares down as much as 25% during Tuesday morning trading, placing the stock on course for its largest one-day decline in decades. The move erased about $70 billion in market value.

IBM issued preliminary second-quarter guidance showing revenue of $17.2 billion and adjusted earnings of $2.93 per share, both below Wall Street expectations. Analysts had forecast revenue of about $17.86 billion and earnings of roughly $3.02 per share. IBM expects second-quarter revenue to increase about 1% year over year, its slowest pace of growth in more than a year.

Industry analysts noted that the company’s results may have a continuing impact across the software sector. “While IBM’s shortfall appears to be very company specific, it may also indicate a crowding out of spend by enterprises,” Gil Luria, Head of Technology Research at D.A. Davidson, told Techstrong.it.

“IBM is the main provider of mainframe hardware and software, which appears to be the impacted category, but it is quite possible companies chose to hold off on those contracts in order to pay for their AI token consumption. If that is the case, we may learn of other non-AI technology companies falling short of expectations over the next few weeks.”

Microsoft, Salesforce, Workday, Autodesk, SAP, Accenture, Cognizant, Infosys and other software and IT services companies also traded lower as investors questioned whether AI infrastructure spending is beginning to squeeze traditional enterprise software purchases.

Sharply Altered Spending Priorities

IBM CEO Arvind Krishna said customers sharply altered spending priorities during the closing weeks of June, directing capital toward AI infrastructure components, including servers, storage systems and memory, as they purchased hardware before anticipated price increases.

Krishna acknowledged IBM did not respond quickly enough to the spending shift and said several large customer contracts failed to close within the quarter, contributing significantly to the revenue shortfall.

The weakness was concentrated in IBM’s software and infrastructure businesses, particularly its mainframe operations.

The results represent one of the clearest indications so far that the AI investment cycle is changing enterprise buying patterns. Rather than expanding overall IT budgets, many companies appear to be shifting spending toward the hardware required to support generative AI.

Cybersecurity has emerged as a competing destination for tech budgets. IBM said enterprises are increasing security investments following advances in AI-based cyber capabilities, including Anthropic’s Mythos model, which has demonstrated a remarkable ability to identify weaknesses in software systems.

IBM has spent years attempting to reduce its dependence on the cyclical mainframe business by expanding higher-margin software operations, including Red Hat, while building partnerships around AI.

The company also highlighted its long-term investment in quantum computing, including plans to invest more than $10 billion toward building a fully functional quantum computer by 2029. Interest in quantum technology has grown after the US government backed efforts to strengthen domestic supply chains.

But those initiatives remain too early to offset weakness in IBM’s core businesses, leaving investors focused on whether enterprise tech budgets will keep shifting toward AI hardware over the coming quarters.

IBM is scheduled to report full second-quarter earnings on July 22.