What IBM’s AI results tell us about enterprise tech spending

On July 14th, 2026, IBM reported that it expects second‑quarter revenue of $17.2 billion, slightly below Wall Street’s $17.86 billion estimate, and pointed to an AI‑driven shift in client spending from software to data‑center infrastructure as a key reason. The announcement sent the company’s stock plunging by 24% and raised the question: what does this say about the future of AI investments?

According to IBM, the issue is not a lack of interest in AI, but a change in where AI‑motivated budgets are going.

“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,” said CEO Arvind Krishna. “While we anticipated some supply‑chain related impact in our expectations, we did not anticipate the magnitude of the capex reprioritization.”

Taken at face value, this quarter is less a verdict on AI itself and more a signal that AI spend is moving into foundations first—which has real implications for how enterprises plan, fund, and measure their own AI programs.

The context behind IBM’s Q2

Gartner predicted that global AI spending will reach a total of $2.59 trillion in 2026, growing at a rate of 47% year-over-year. According to John-David Lovelock, Distinguished VP Analyst at Gartner, AI infrastructure such as AI-optimized servers, network fabric, and semiconductors and devices would be the largest segment of the market, accounting for over 45% of this total spend.

“Within this segment,” explained Lovelock, “spending on AI-optimized servers will triple over the next five years to become the largest subsegment, as cloud services providers expand capacity in anticipation of the workloads created by GenAI models and agentic workflows.”

By comparison, worldwide spending for AI services is expected to reach $585,527 million this year, followed by AI software spend at $453,209 million. While these investments are nothing to scoff at, they still sit behind the surge going toward infrastructure. 

Companies like Micron, SK hynix, and Samsung are thriving, as more of their production is pointed at AI‑grade parts instead of general‑purpose components. Over the past year they’ve reallocated lines toward high‑bandwidth memory and other specialized chips for AI data centers, and much of that capacity is already spoken for under long‑term contracts. 

"This is an ugly moment for IBM and software stocks...the big question will be how long the shift to infrastructure and cybersecurity lasts," said Chris Beauchamp, chief market ​analyst at IG Group.

With that said, however, IBM has its own infrastructure division, which saw revenue go down by 7%. This can be chalked up to IBM’s mainframe‑heavy infrastructure business and to timing: late in the quarter, clients shifted capex away from IBM software and mainframe contracts into other vendors’ servers, storage, and memory.

"This quarter we faltered," said Krishna. “We did not adapt and move quickly enough, and numerous large deals failed to close on the timelines we expected, driving the majority of our shortfall."

Is AI software in danger?

The short answer is no. While AI software investment is behind hardware, that gap has more to do with sequencing than with a loss of confidence in software. 

Buyers are spending this phase on capacity: GPUs, memory, data centers, and the core infrastructure they need before they can run anything serious at scale. It might slow some platform and application decisions by a few quarters, but once this infrastructure is in place, companies will still require software to fully utilize it. 

IBM itself framed this as an execution and timing problem, with deals slipping out of the quarter rather than customers walking away from AI‑related software altogether. Additionally, the software side of the company’s business still grew, indicating that customers are continuing to fund AI‑related projects even as more budget is temporarily tied up in infrastructure.

What does IBM’s announcement mean for enterprises?

IBM’s quarter is a reminder that AI spend now lives inside the same budgets as everything else. When more money goes into AI‑related infrastructure and platforms, something else will slow down or slip, and in this case it was a mix of software, consulting, and mainframe‑centric infrastructure deals that didn’t land in time.

For enterprises, the question is not whether to spend on AI, but in what order. If you choose to bring forward spending on cloud capacity, data‑center contracts, or managed AI services, you should expect some application work, integration projects, or process changes to move into later quarters. That is a defensible trade‑off, as long as it’s visible in your roadmap and consistent with what you’re telling your board about when AI will start to show up in how the business actually runs.

The other implication is that “AI investment” now has to be explained in more detail than a single number. Hardware, software, and services can move in different directions inside the same plan. When you talk about your own approach, it helps to be clear about which part you’re accelerating, which part you’re deferring, and how those pieces are supposed to come together over the next 12–24 months.

Planning your next AI budget cycle

IBM’s results highlight a reality many enterprises are already feeling: AI is no longer an experimental line item—it’s competing directly with other technology priorities inside the same budget. When infrastructure spending moves first, software, integration, and change management often get pushed to later quarters, which can make it harder to realize the value you expected from those AI investments.

To stay ahead of that pattern, it helps to treat AI planning as an end‑to‑end exercise rather than a series of disconnected purchases. That means sequencing infrastructure and software work deliberately, setting clear expectations about timing and impact, and defining what “good” looks like over the next 12–24 months in terms of productivity, customer experience, and risk.

At FullStack Labs, we partner with companies that want to take this more structured approach to AI. We help teams align budgets with realistic roadmaps, design solutions that fit their data and industry, and track progress against measurable outcomes instead of hype. If you’re looking to turn AI from a cost line into a durable source of value, our consultants and engineers can work with you to plan, build, and refine the systems that make that possible.

Contact us today if you’re interested in working with us on your next AI project.