Enterprise Technology Investment Is Rising—But Can Organizations Turn AI Spending Into Value?

Enterprise Technology Investment Is Rising—But Can Organizations Turn AI Spending Into Value?

AI is testing how companies budget, deploy infrastructure, and govern technology. The next challenge is making sure investment and productivity gains translate into real organizational capability.

Enterprise technology investment accelerated again in October, reaching its second-highest level in three years, according to the November Board Monthly Update from Esteban Kolsky of Constellation Research. Real technology investment grew 20.6% year over year, comparing Q3 with the same quarter a year earlier, with spending directed primarily toward infrastructure and, to a lesser extent, services.

But increased spending does not automatically mean increased value. As organizations implement AI, they are being pushed to reassess where technology budgets go, what results they expect, and how quickly they can change course when evidence shifts.

AI puts dynamic budgeting to the test

AI is becoming a proving ground for dynamic budgeting: the ability to adjust investment as needs and results change. Organizations are moving away from broad, loosely defined AI spending and allocating funds more deliberately across infrastructure, software, services, data science, and custom models.

The value case remains unsettled. Only 13% of CEOs report significant or transformational value from their AI implementations, even as investment continues to grow. That reflects expectations that AI will matter—but also raises a practical question: What should organizations do when an initiative is not delivering?

Many companies still lack clear rules for stopping or redirecting funding from underperforming projects. Without those rules, organizations risk continuing to invest simply because a project is already underway. Dynamic budgeting requires more than reallocating money; it requires a willingness to act when the evidence changes.

Productivity gains need a destination

AI can automate tasks and release capacity, but that capacity does not automatically become innovation. In the update, only 21% of released capacity is being redirected toward innovative work. Much of the rest is going unused.

The opportunity is to treat released time and resources as something to actively manage. Organizations can ask how that capacity might support growth, improve service, or make the business more resilient. A meaningful measure of AI-driven productivity, then, is not only how much work is automated—it is what the organization is able to do with the capacity it gains.

Infrastructure is essential—and increasingly constrained

Infrastructure is becoming both more important and more challenging to manage. More than 62% of organizations report double-digit price increases in recent technology purchases, including computing from hyperscalers and software offerings priced around agents or outcomes.

At the same time, financial pressure and outages among infrastructure providers highlight the risks of concentrating technology spending. In response, CFOs are demanding greater cost discipline and clearer evidence for how proposed investments will be used.

Organizations have more technology choices, but choice alone does not create resilience. Resilience depends on whether workloads and resources can be redirected when conditions change—and whether the organization can make those adjustments quickly enough.

Governance and cyber readiness must catch up

The gap between planned agentic AI adoption and governance readiness is significant. 71% of organizations expect to deploy agentic AI within the next 12 months, while only 21% report having the governance capabilities to manage those deployments.

Cybersecurity plans face a similar challenge. While 84% of CIOs expect cyber budgets to increase, only 40% have a fully operational cyber continuity plan. Fewer than 10% include agents in those plans.

More spending and more deployment do not, by themselves, provide the ability to govern, recover, or respond. As AI systems become part of business operations, organizations need practical ways to manage their authority and adjust—or stop—their actions when operating conditions change.

What to watch next

The update identifies four questions for the months ahead:

1. Can organizations move capital when the evidence changes? If a project is not working, can funding be redirected to a better alternative?
2. Can productivity gains become new capabilities? Will released human, technology, and operational capacity support growth, service, or resilience?
3. Are practical alternatives viable today? Custom models and edge inference deserve attention alongside large-scale infrastructure investments.
4. Can deployed systems adjust to changing conditions? Organizations need to know whether they can demonstrate machine authority and change or stop system behavior when needed.

From investment to adaptability

The central challenge is not simply how much organizations spend on technology or AI. It is whether they can connect that spending to useful outcomes—and adapt when projects, costs, or operating conditions change.

That means setting clearer rules for investment decisions, giving released capacity a deliberate purpose, assessing infrastructure risk, and building governance and continuity capabilities alongside deployment. As AI adoption grows, the organizations best positioned to realize value will be those that can learn, reallocate, and respond—not just invest.