IBM has lowered its annual revenue growth forecast after discovering that the artificial intelligence spending boom does not benefit every part of the technology market equally.
Corporate customers are still spending. Quite heavily, in some cases. The problem for IBM is where that money is going.
Businesses have started prioritizing AI-focused servers, storage systems and increasingly expensive memory components. That shift has squeezed budgets for other technology purchases, including some of the software and mainframe products that remain important to IBM’s business.
It is an awkward position. IBM has spent years presenting itself as an AI and hybrid cloud company, yet the rush to build AI infrastructure has disrupted its own sales cycle.
IBM Lowers Its 2026 Revenue Growth Outlook
IBM now expects revenue to grow between 4% and 5% at constant currency during 2026. The company previously projected growth above 5%.
That is not a collapse, but it is still a meaningful downgrade. Investors had expected IBM’s software portfolio, consulting operations and AI products to keep the company comfortably above its earlier target.
IBM kept one important part of its outlook unchanged. The company still expects full-year free cash flow to increase by around $1 billion compared with 2025.
Management is clearly trying to draw a line between weaker revenue growth and the overall health of the business. Cash generation remains solid. Revenue momentum, however, has become harder to predict.
Customers Are Moving Budgets Toward AI Hardware
The explanation is not that companies suddenly stopped investing in technology.
They changed the shopping list.
IBM CEO Arvind Krishna said customers shifted part of their quarterly capital spending toward servers, storage and memory during the final weeks of June. Businesses wanted to secure supply-constrained equipment before further price increases arrived.
Those purchases are closely connected to the wider AI infrastructure race. Training and operating large AI systems requires powerful servers, large amounts of storage and specialized memory. None of that comes cheaply.
When infrastructure costs rise, something else often gets delayed.
For IBM, that meant several expected software and mainframe-related transactions did not close before the quarter ended. The company also said rapidly changing cybersecurity concerns distracted some customers and further complicated purchasing decisions.
IBM Admits It Failed to React Quickly Enough
IBM did not blame the entire shortfall on customers or supply chains.
Krishna acknowledged that the company’s own execution fell short. Sales teams did not adjust quickly enough when buying priorities changed, and several large deals slipped beyond their expected closing dates.
That admission matters.
Enterprise technology sales are rarely simple. Contracts can take months to negotiate, involve multiple departments and disappear from one quarter’s results because of a relatively small delay. Even so, IBM had expected many of those transactions to close.
They did not.
The company now plans to expand sales coverage across thousands of additional clients. IBM is also investing in more specialized technical staff, including forward-deployed engineers who can work more closely with customers during large AI deployments.
Second-Quarter Revenue Reaches $17.2 Billion
IBM generated $17.2 billion in revenue during the second quarter of 2026, an increase of 1% from the same period last year.
Net income came in at approximately $2.2 billion. Adjusted earnings reached $2.93 per share, slightly below the average analyst estimate of $2.97 reported by Reuters.
The headline number looked stable. The individual business units told a messier story.
Software revenue increased 5% to $7.8 billion, supported by stronger results from Red Hat and IBM’s data portfolio. Consulting revenue remained broadly flat at $5.3 billion.
Infrastructure revenue dropped 7% to $3.8 billion.
That decline carried most of the damage.
IBM Mainframe Revenue Drops Sharply
IBM Z mainframe revenue fell 42% during the quarter.
A large decline was always likely because IBM was moving beyond the early launch period of its z17 mainframe system. New product cycles usually produce a surge in initial orders, followed by slower comparisons later.
Still, IBM expected the slowdown to be milder.
Transaction Processing revenue, which includes software closely connected to mainframe workloads, also fell 8%. The weakness suggests that slower hardware purchasing affected more than the machines themselves.
Mainframes remain deeply embedded in banks, government agencies, airlines and other organizations that process enormous volumes of critical transactions. They are not disappearing overnight.
Quarterly demand can be uneven, though, especially when the same customers are being asked to fund expensive AI projects at the same time.
Not Every Part of IBM’s Infrastructure Business Struggled
There was a strange contrast inside IBM’s infrastructure division.
While mainframe revenue sank, distributed infrastructure revenue climbed 37%. Demand for IBM Power systems and storage products reached what the company described as a record pace.
IBM ended the quarter with a distributed infrastructure order backlog of almost $500 million.
That supports the company’s argument that customers are not abandoning IBM entirely. They are buying the products that fit their immediate infrastructure priorities and delaying others.
Red Hat revenue also increased 11%, while IBM’s data business grew 19%. Recent acquisitions, including HashiCorp and Confluent, delivered stronger performances as well.
There are healthy pieces here. They simply could not erase the mainframe slowdown and missed enterprise deals.
AI Spending Is Creating Winners and Delays at the Same Time
The wider lesson is uncomfortable for established technology companies.
AI spending can lift demand across the industry while creating pressure elsewhere. A business may have the same technology budget it had three months ago, but suddenly a much larger share must cover GPUs, memory, networking equipment, storage and data-center capacity.
Traditional software upgrades can wait. A mainframe expansion may move into the next quarter. A consulting project gets reviewed again.
That does not mean companies have lost interest in AI software. It means the infrastructure required to operate AI systems is swallowing a larger portion of near-term budgets.
IBM is experiencing that shift from both sides. It sells AI software and infrastructure, yet it also depends on customers continuing to buy established products while they finance their AI transitions.
IBM Is Still Betting on AI, Automation and Quantum Computing
Despite the weaker forecast, IBM is not pulling back from its long-term technology strategy.
The company said it will continue investing in Red Hat, watsonx, automation and enterprise AI services. It also plans to use AI internally to improve software development, sales productivity and supply-chain operations.
Quantum computing remains another large commitment. IBM recently announced plans to invest more than $10 billion in quantum technology over five years and says it remains on track to deliver a large-scale fault-tolerant quantum computer by 2029.
These investments sound ambitious next to a lowered revenue outlook. That tension is likely to remain.
IBM needs to keep funding the technologies that could drive its next phase of growth while proving that its existing businesses can handle a sudden change in customer spending.
IBM’s Next Test Is Execution
The reduced IBM revenue growth forecast is not simply a warning that enterprise technology demand is weakening.
Demand is moving.
Customers are prioritizing the equipment needed to build AI capacity, even when that means postponing software, mainframe or consulting purchases. IBM saw that shift happening but failed to respond quickly enough before the quarter closed.
Now the company has to show that delayed deals were actually delayed, not lost.
The next few quarters should make that clearer. Stronger software growth, continued demand for distributed infrastructure and better sales execution could make the second quarter look like an expensive timing problem.
Another round of missed deals would suggest something more structural.
Sources
- Reuters: IBM cuts yearly revenue growth forecast as customers prioritize AI infrastructure spending
- IBM: Second-Quarter 2026 Results
- IBM: Arvind Krishna’s Letter to Investors
- IBM Investor Relations: Second-Quarter 2026 Earnings