AI software budgets are under serious pressure worldwide, and IBM just gave the clearest proof yet. IBM reported adjusted earnings of $2.93 per share on revenue of $17.2 billion, missing analysts’ expectations of earnings of $3.01 per share and revenue of $17.86 billion. The reason was not weak demand for technology. It was a fast, sharp move by big companies to spend on AI hardware instead of software and services.
What Happened at IBM in Q2 2026
Chief Executive Arvind Krishna attributed the disappointing quarter to an unexpected shift in customer spending toward AI-related hardware purchases, saying: “In the last few weeks of June, we saw clients shift their quarterly capex spend toward servers, storage, and memory purchases to secure supply-constrained infrastructure ahead of expected price increases.”
Krishna added that the company had “faltered” in adapting quickly enough and that “numerous large deals” had failed to close as expected. The stock reacted fast. IBM’s shares went down 26% in early trading, putting the stock on track for an even steeper single-day decline than it suffered during the 1987 “Black Monday” crash.
The sell-off spread across the software sector, affecting companies including Microsoft, Salesforce, ServiceNow, and Intuit. Goldman Sachs went further. Goldman Sachs warned that the IBM incident would validate a “software bear market scenario,” with widespread selling expected in the software and services sector.
Why AI Software Budgets Are Being Squeezed
The core problem is simple: companies have fixed IT budgets, and AI infrastructure is eating a bigger slice of that fixed pie.
Organizations building AI systems need accelerators, networking equipment, storage capacity, data-center resources, and supporting infrastructure. Every dollar directed toward that foundation can leave less funding available for software deployments, consulting projects, and modernization initiatives.
The warning highlights a growing divide in the technology industry. While companies building AI infrastructure, including chipmakers, cloud providers and data center operators, continue to benefit from heavy investment, software companies are increasingly being forced to prove that AI will expand rather than replace their businesses.
AI consumption does not necessarily grow predictably or linearly; unlike fixed SaaS licensing, token-based usage can fluctuate dramatically depending on user behavior, workflows, and model selection. “The rapid acceleration of token spend is likely already outpacing forecasts made during this year’s budget cycle,” according to one enterprise AI analyst.
Big Tech is expected to spend more than $700 billion on AI infrastructure this year. That spending rush is pushing up prices for servers and memory, which is why enterprise buyers are rushing to secure supply before costs rise further. The result: AI software budgets for things like ERP upgrades, licensing renewals, and consulting projects get pushed back.
How Bad Was IBM’s Segment Performance
IBM said software revenue increased 5% during the quarter, consulting revenue was broadly flat rising 1% at constant currency, while infrastructure revenue declined 7%. But context matters. IBM had reported first-quarter revenue growth of 9%, with software up 11% and infrastructure up 15%. The preliminary second-quarter figures show growth slowing to 1%, 5%, and negative 7% respectively, a 22-percentage-point swing in infrastructure.
IBM is trying to reassure investors by pointing to longer plans. IBM highlighted its heavy investments in quantum computing including more than $10 billion to build the first large-scale quantum computer by 2029. But IBM’s quantum efforts and expanding its AI partnerships, including with OpenAI, are still in early stages and are not yet large enough to materially offset weakness in its core software and infrastructure businesses. Full Q2 results are due on 22 July.
Why This Matters for Pakistani Banks, Telcos, and Enterprises
Pakistani companies may not buy IBM mainframes directly, but the global budget shift described above is the same one now playing out inside every large Pakistani bank, telco, and enterprise that is planning its IT spending for 2026 and beyond.
Pakistan’s enterprise software development market reached $850 million in 2026, driven by digital transformation, custom ERP development, banking software modernization, and healthcare IT solutions. That is a big number for a developing market, and it means local decision-makers are facing exactly the same dilemma as IBM’s global clients: do we spend on AI infrastructure now, or keep paying for existing software?
Pakistani banks are under pressure from the State Bank of Pakistan to digitize faster. Telcos are building data infrastructure. Government departments are pushing digital services. All of these organisations must now decide whether to keep renewing traditional software or divert rupees toward AI compute, cloud servers, and storage. IBM’s warning is a signal that even large global companies got this trade-off wrong.
Industry observers say the global shift toward cloud computing, AI-powered business operations, cybersecurity demand, and remote digital services has significantly increased outsourcing opportunities for emerging technology markets such as Pakistan. That is the upside. But the downside is that local IT buyers may defer long-planned software upgrades or ERP rollouts if AI hardware costs eat into their budgets first.
Enterprises are entering a new phase of the AI cycle, one defined less by experimentation and more by trade-offs, prioritization, and budget discipline. Pakistani CIOs and CFOs need to take this seriously. Rushing into AI hardware without a clear plan could mean critical legacy software goes un-upgraded, creating risk in banking transaction systems and telecom billing platforms that millions of Pakistanis rely on every day.
What Pakistan’s Own AI Ambitions Mean in This Context
Pakistan is not a passive observer here. In July 2025, Pakistan’s federal cabinet approved the National AI Policy 2025, which includes an AI Council, a master plan, AI Innovation and Venture Funds, and a headline target to train one million AI professionals. The policy is ambitious. But the three structural challenges facing Pakistan’s AI ecosystem are infrastructure (unreliable electricity and limited compute capacity), talent (only 10% of the IT workforce currently holds AI skills), and data access.
This means Pakistani enterprises face a double squeeze: global AI hardware prices are rising, and local compute capacity is limited. The IBM story is a reminder that simply wanting to build AI is not enough. Budget discipline matters more now than at any point in the last decade.
“The real challenge is rarely just buying the technology,” one expert noted. “It’s redesigning workflows, training teams, managing risk, and ensuring the AI meaningfully improves operational outcomes at scale.” That advice applies equally in Karachi, Lahore, or New York.
For Pakistani IT managers, the practical read is this: before locking in new AI hardware or cloud commitments, map out exactly which software contracts you are planning to defer. IBM’s clients delayed those decisions and caught their own company off guard. Local enterprises cannot afford the same mistake.
Frequently Asked Questions
Why did IBM miss its Q2 2026 revenue target?
The shortfall was primarily due to enterprise customers shifting capital expenditures from traditional software to AI infrastructure such as servers and storage, resulting in several large contracts remaining unsigned. IBM CEO Arvind Krishna attributed the miss to a “sudden shift” in customer capital expenditure priorities amid the AI boom driving high demand for servers, storage, and memory.
Are AI software budgets being cut globally?
They are not being cut exactly, but they are being redirected. The shortfall at IBM was not due to shrinking AI budgets, but a reallocation: clients prioritized acquiring compute, storage, and memory amid supply constraints and expected price increases, delaying software deals. The total tech spend is still growing, but software gets a smaller share of it.
How does this affect Pakistan’s enterprise software market?
For technology leaders, the result is a more complicated purchasing environment in which AI infrastructure, cyber resilience, and traditional software must compete for the same capital. Pakistani banks, telcos, and government bodies face the same choice. Spending more on servers and AI compute this year may mean deferring ERP upgrades and software licences.
Will this trend last, or is it a short-term shift?
IBM will release its complete second-quarter results on July 22, providing a clearer view of whether this spending shift represents a temporary procurement cycle or a broader change in enterprise technology strategy. Analysts remain divided on whether this shift is temporary or long-term. Most experts say that as AI hardware supply catches up, the pressure on software budgets should ease, but the discipline of tighter budget management is likely here to stay.