Companies now invest more in computing for Al workloads than in the people running them.

getty

The cost of the workforce has driven business budgets for decades—salaries, benefits, and retention built on managing human capital. But a key inflection point merits discussion for those tracking enterprise IT budget trends.

Now, companies invest more in computing for AI workloads than in the people running them. This shift is significant.

"The costs of compute have exceeded the costs of my people," said Bryan Catanzaro, VP Applied Deep Learning at Nvidia. This isn’t just Catanzaro’s view—it's an expanding trend, especially for AI-focused businesses.

Uber’s CTO spent its entire 2026 AI budget before March.

Swan AI CEO Amos Bar-Joseph announced on LinkedIn that they used their whole AI budget in two months, praising the autonomy: "...building the first autonomous business—scaling with intelligence, not headcount."

The key takeaway: AI compute costs are reshaping enterprise investment priorities and redefining IT value assessment.

But let's dig a little deeper into the implications of this trend.

According to the research firm Gartner, total worldwide IT spending is set to hit a record high of $6.31 trillion in 2026 – up 13.5% from 2025.

Cloud and AI software subscriptions, along with AI infrastructure and services, drive most of this growth—even as some IT managers slow spending.

Investing more in AI than in people is great if it translates into tangible results—as with all other major technologies over the past four decades or so. Otherwise, it's just another tech bubble.

The proof will be in the quarterly earnings call when shareholders start asking for ROI on that investment. And that could mean a number of things – savings in operations, increased efficiency, better time-to-market, higher per-head output rates, or even the creation of new revenue lines from AI-based products and services that would never have existed otherwise.

Businesses will debate this question for years. Answers will vary widely by industry, use case, and organization. But let me throw another variable into the equation.

Early adopters may pay more as AI labs raise prices as adoption grows. And that's why investors in OpenAI are keeping an eye on their competition, Anthropic. At least one of them believes Codex outperforms the competition, Claude Code, and that makes a significant difference when computing costs are already soaring.

It's all well and good to spend more money on AI now, but that might bite companies back later when the cost of AI starts to outweigh potential savings – or other value-driven benefits.

I've seen enough cycles in technology in the past decades—and they usually followed a similar pattern: a breakthrough in capabilities, advantages to early adopters, increased expenditures, and then—due to rising costs and increasing hype—a sobering reality check in the form of a more rigorous conversation about value generation.

The central takeaway now: AI’s value isn't just spending more—it's tracking ROI and adapting to new economic conditions.

Human capital and AI compute aren’t mutually exclusive. Spending more on AI isn’t a strategy. The companies that prevail won’t be the ones with the largest AI budgets, but the ones that can clearly tie that spending to measurable outcomes. Success depends on treating AI not as a signal of innovation, but as an investment measuring business value and adapting to changing economics.