CIOs are not uniformly turning away from AI. Some are accelerating deployment while others are staging it because accountability, security controls, costs, and evidence of business value are not always keeping pace. The quiet constraint is operational: leaders may be responsible for AI systems they cannot fully see or control.
Why accountability can outpace control
As AI moves from experiments into business processes, CIOs are expected to oversee systems that may be adopted across departments, connected to company data, or used through tools outside formal IT approval. IBM’s June 2026 study found that two-thirds of surveyed CIOs and CTOs said they were accountable for AI systems they did not fully control (IBM Institute for Business Value).
As an Amazon Associate I earn from qualifying purchases.
That mismatch makes visibility a practical prerequisite for expansion. An organization cannot reliably assess risk, assign responsibility, or respond to an incident if it does not know which systems are in use, what data they can access, and who owns them. Roush CIO Chris Pesola described the goal as creating visibility and partnership rather than eliminating shadow IT: “The goal isn’t to eliminate shadow IT—it’s to create visibility and a partnership, so teams can get help when they need it without slowing down.” The statement appeared in IBM’s June 2026 release, not as a survey result.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Why security and governance slow scaling
AI agents can take actions across systems, making security and compliance questions more consequential than they are for a standalone assistant. IBM reported that 59% of surveyed technology executives named security and compliance among the top barriers to scaling AI agents (IBM, 2026). That figure describes respondents in IBM’s study; it is not a universal measure of CIO opinion.
#1 Best Overall
Governance can be handled through manual review, but review alone may become a bottleneck as the number of tools and use cases grows. IBM’s study analysis reported 25% fewer incidents in organizations embedding controls into AI systems than in organizations relying on manual governance. That is a reported comparison, not a guaranteed causal result for every company. The operational implication is that controls built into systems can make oversight more scalable than relying only on after-the-fact approval.
Why more AI activity does not guarantee business value
Deploying pilots or adding AI features is not the same as meeting a business objective. In CIO.com’s 2026 State of the CIO survey, which canvassed 662 IT leaders and 249 line-of-business users, 19% of respondents said AI initiatives met or exceeded business goals; 18% said fewer than one-third of AI use cases met defined expectations (CIO.com). These are survey responses, not a universal success rate or an independent audit of every initiative.
Rank #2
The gap between activity and results gives CIOs reason to ask which use cases solve a sufficiently important problem, how success will be measured, and whether benefits justify operational and financial costs. Older corroborating evidence from Salesforce’s 2024 CIO research found that 68% of surveyed CIOs believed business partners had unreasonable expectations about when AI would produce ROI (Salesforce). That finding reflects respondents’ perceptions in 2024, not a current forecast of returns.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Why approval processes matter
Formal approval is one way to establish ownership and review risks before a system reaches wider use. In CIO.com’s 2026 survey, 53% of respondents said their organization had established an official AI approval process. Approval alone does not establish that a system is safe or effective, but a defined process can clarify who evaluates a proposed use, what information it may handle, and how its results will be monitored.
Rank #3
Why cost scrutiny does not mean a universal slowdown
Token costs can affect the economics of frequent or large-scale AI use, but the available survey findings do not support a simple claim that CIOs broadly are slowing deployments. EY’s July 2026 survey reported that 15% of surveyed AI-investing senior leaders were slowing rollout in response to token costs, while 29% were speeding it up (EY). The responses show differing decisions among surveyed leaders, not a single trend shared by all organizations.
At the same time, IBM projected that AI’s share of IT budgets would rise from just under 15% in 2025 to nearly 25% by 2027 (IBM, 2026 PDF). That projection concerns budget allocation, while EY’s findings concern reported rollout pace in response to token costs. They are different measures, but together illustrate how investment can grow while organizations scrutinize where and how quickly they deploy.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What a responsible pace looks like
Taking more time need not mean stopping adoption. A practical approach is to expand use where the business case and operating controls are clear, while improving oversight before higher-risk systems are given broader access or autonomy.
- Make ownership visible: identify the business owner, technical owner, and accountable approver for each significant AI use.
- Know what is in use: track sanctioned systems and create a route for employees to disclose or request help with tools already being used.
- Build controls into deployment: define data access, permissions, review points, and incident handling as part of the system rather than relying only on manual governance.
- Measure outcomes: set a business goal and a way to assess it before treating adoption or pilot counts as evidence of value.
- Match rollout to economics and risk: evaluate usage costs and the consequences of errors for each use case instead of assuming every AI application warrants the same pace.
The surveys cited here come from different publishers, respondent groups, years, and question wording, and their samples are not directly comparable. They support a qualified conclusion: control, security, ROI expectations, and costs can lead some leaders to stage or reconsider rollout, even as others continue to accelerate investment.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
