CIOs Face Mounting Pressure as AI Spending Surges Beyond IT Budgets
AI is no longer an IT line item; it is becoming a distributed operating expense embedded across functions. That changes the CIO’s job from budget owner to enterprise orchestrator. When most spend sits outside central IT, the management challenge is not simply cost control but establishing decision rights, architecture guardrails, and a common method for evaluating use cases before fragmented commitments harden into long-term complexity.
The bigger risk is a widening gap between experimentation volume and enterprise value. If business units fund pilots independently while ROI measures remain inconsistent, leaders can end up with duplicated models, overlapping vendors, unplanned inference costs, and unclear accountability for outcomes. CIOs should push for an AI investment taxonomy that separates foundational platform spend, functional use cases, and transformational process redesign, with different hurdle rates and review cadences for each.
The article also signals that benefits realization will depend less on model performance than on operating model change. If most value comes from workflow redesign and adoption, then governance cannot stop at model approval. It must extend to process ownership, workforce implications, control design, and post-launch measurement. This is where partnership with CFOs and business leaders becomes critical: not just to approve funding, but to define what counts as value, when to stop underperforming initiatives, and where to keep investing despite delayed payback.
Practical next questions for IT leaders:
- Who has authority to approve AI spend, and who owns enterprise visibility across it?
- Which use cases justify shared platforms versus local experimentation?
- Are inference, security, and integration costs visible in business cases?
- What adoption, process, and risk metrics determine whether pilots scale, pause, or stop?
Enterprise technology leaders find themselves in a tightening spot. AI investments have more than doubled as a share of revenue in the past year. Yet measurable returns remain elusive for many. And the bulk of that money now flows outside traditional IT departments.
Budgets Shift as AI Moves Into Business Operations
Spending on artificial intelligence jumped from 1.7% of revenue in late 2025 to 3.3% today, according to a new report from BCG. The research group surveyed more than 1,300 CIOs and other tech leaders. Eighty percent of AI spending now sits outside the IT budget. (https://www.ciodive.com/news/ai-spending-spreading-beyond-it/831823/)
That dispersion makes sense on paper. AI has become business transformation. Yet it leaves CIOs chasing visibility. Vlad Lukic, global leader for BCG’s tech and digital advantage practice, put it plainly in an email to CIO Dive. “What CIOs do need, though, is visibility and a common management model across that spend.” (https://www.ciodive.com/news/ai-spending-spreading-beyond-it/831823/)
But. The picture grows more complicated. Worldwide AI spending will increase 49.5% in 2026 to $2.7 trillion, Gartner predicts. Growth continues at 36.2% the following year. John-David Lovelock, distinguished VP analyst at Gartner, noted that CIOs received some net new money for AI in 2024 and 2025. “So there wasn’t a diversion, and now more of their spending is about rebranding than diversion.” (https://www.cio.com/article/4225785/ai-spend-will-jump-49-5-in-2026-says-gartner.html)
Infrastructure spending tells its own story. Gartner expects AI-optimized infrastructure-as-a-service spend to nearly double by year-end, hitting $42 billion. Inference spending this year — $23.3 billion — already tops training at $19 billion. “Now, enterprises are embedding AI into applications, business processes and customer experiences, which requires continuous inference rather than periodic training,” one Gartner analysis stated. (https://www.ciodive.com/news/AI-spending-soars-enterprise-maturity/827488/)
Executives talk a good game about productivity. Reality bites harder. Seventy-two percent of organizations have scaled less than a quarter of their AI pilots successfully. Two-thirds struggle to measure ROI. The Infosys report that surfaced these numbers also found nearly three-quarters say the rush to prove short-term returns inhibits longer-term experimentation. (https://www.ciodive.com/news/CIOs-waiting-AI-savings-ROI/828883/)
So organizations keep pouring money in. Forrester projected global technology spending to reach $5.6 trillion in 2026. Hyperscalers plan more than $500 billion in AI-related capital expenditures this year alone. The buildout ranks among the largest infrastructure projects in history.
Agentic systems add another layer. These autonomous tools promise to reshape entire workflows rather than handle single tasks. BCG expects them to drive two-fifths of all AI value by 2030. Forty-two percent of companies plan to grant agents autonomy by then. Lukic highlighted the shift. Companies report real value where agents overhaul end-to-end processes. (https://www.ciodive.com/news/ai-spending-spreading-beyond-it/831823/)
The recommended split feels telling. BCG suggests organizations dedicate just 10% of effort to algorithms, 20% to technology and data, and 70% to people, organization and processes. A strong tech stack delivers little in a static workflow or with reluctant employees.
CIOs also wrestle with governance gaps. Recent surveys show many have lost oversight of AI agents created by employees. Budgets face scrutiny. If performance targets slip by the end of 2026, cuts or freezes loom. Reputation and career trajectories now tie directly to AI outcomes.
Security leaders feel parallel pressure. Governance tops concerns as organizations mix public, open-weight and private AI models. Limited visibility into AI-driven traffic hampers efforts. Eighty percent cite inadequate controls around unsanctioned AI use as their top data security challenge. (https://www.gigamon.com/company/news-and-events/newsroom/2026-survey-shift-to-hybrid-ai-drives-governance-to-top-of-the-ciso-agenda.html)
Proofpoint’s 2026 Voice of the CISO report captured the mood. Seventy-eight percent of CISOs now view generative AI as a security risk, up 18 points from the prior year. Eighty-five percent prioritize safe AI use over the next two years. Yet 79% must manage those risks without added resources. (https://www.proofpoint.com/us/2026-voice-ciso)
IBM’s latest study adds optimism on timing. Nearly eight in 10 executives believe 2030 will mark the year AI delivers significant revenue impact. Only 40% say it boosts revenues today. Investment as a share of revenue will more than double over the next four years. Integration issues worry 68% of leaders. (https://newsroom.ibm.com/2026-09-30-ibm-study-as-ai-scales-enterprise-wide,-cfos-play-an-expanded-role-in-transformation)
CFOs step into bigger roles too. Sixty-two percent report expanded responsibility in technology or AI strategy. More than half expect by 2030 to help design financial and ethical guardrails for the technology. Organizations led by AI-first CFOs already post 23% higher revenue growth rates.
The Fortune CIO Intelligence briefing captured similar tensions on September 30. Technology chiefs balance aggressive deployment with the need for measurable outcomes and proper oversight. (https://fortune.com/2026/09/30/cio-intelligence-sept-30/)
Predictions for 2026 point to smaller development teams thanks to AI-native platforms. Purpose-built computing gains traction. Workforce strategies require fresh thinking as data roles potentially shrink. Governance practices tighten. Forrester expects 60% of Fortune 100 companies to appoint a head of AI governance next year.
Costs complicate matters further. Models grow more token-hungry faster than prices fall. Token expenses widen. Any efficiency gains risk being swallowed by rising consumption. The cautious era has arrived. All-gas-no-brakes enthusiasm gives way to pragmatic questions about value, risk and sustainability.
CIOs who secure transparent views across spending, tie investments to clear business outcomes, and build cross-functional governance stand the best chance. The technology won’t wait. Neither will boards or shareholders. Success now hinges on turning scattered experiments into coordinated, measurable transformation.
Enjoyed this article? Sign up for our newsletter to receive regular insights and stay connected.