For years, enterprises dealt with employees using unauthorized tools because official systems were too slow. Agentic artificial intelligence (AI) changes the problem. These tools can trigger tasks, influence workflows and touch internal systems without a clear owner, audit trail or validation point. At that point, Shadow IT becomes shadow operations.
When something goes wrong, it may not look like an AI problem. It may appear as a cost overrun, a compliance gap or an operational issue that no one can easily trace back.
That changes the mandate for executive leaders. The question is no longer about restricting AI access. It is whether your organization can govern autonomous systems before they become part of how the business runs.
The readiness gap
The business ambition is already clear, but organizational readiness to govern and support that ambition is less certain.
Nutanix’s latest Enterprise Cloud Index (ECI) Report quantifies this gap. Across Asia Pacific, between 77 and 87 percent of enterprises say their infrastructure is not ready for AI workloads.
In Singapore, the gap is sharper. Recent research revealed that 72 percent of local business leaders plan to deploy agentic AI within two years, up from 15 percent today. At the same time, 93 percent of Singapore organizations say silos between business units and IT make effective technology execution difficult, compared with 82 percent globally.
When ambition moves faster than the governed path, teams find workarounds. That is more than an inconvenience. It shows the enterprise operating model has not kept pace with how business value is created today. The risk often stays outside IT’s line of sight until those workflows touch enterprise systems.
From shadow tools to shadow operations
Shadow IT was the first version of this pattern. Employees reached for unauthorized applications because official systems could not move fast enough. The same dynamic is now playing out with AI, only the stakes are higher. Our ECI report found that 79 percent of IT executives have encountered AI applications or agents implemented by employees outside IT functions.
What makes this phase different is that AI has moved from being accessed outside enterprise control to acting beyond it. Shadow operations emerge when AI tools execute tasks or trigger actions across enterprise systems without clear ownership, validation or executive oversight.
OpenClaw, an open-source AI agent, is a clear example of how quickly this pattern can develop. Its adoption has been driven largely by individual experimentation rather than formal IT integration. For IT leaders, the concern is the exposure created when agentic tools operate without clear enterprise controls.
That exposure is not simply a user-behavior problem. The infrastructure and governance frameworks simply haven’t kept pace with how the business now operates.
Make governance useful
The answer is not to slow experimentation. The real mandate for leadership is to make the governed path the easier path for teams to use.
The first step is operational visibility. That means knowing which AI tools are in use, what data they access, which workflows they influence and who owns them from a business perspective. Without that baseline, risk management remains theoretical.
Governance fails when business teams must choose between speed and safety. Organisations need a trusted platform that balances innovation with control. Access controls, monitoring, policy enforcement and clear ownership must be built in from the start. Every agentic workflow needs a clear owner, an access boundary, an audit trail and a validation point before it touches live production systems.
Modernize for Control, Not Restriction
Those controls depend on hybrid cloud infrastructure that can apply policy, manage risk and scale resources consistently across environments.
AI workloads are continuous, data-sensitive and rarely confined to one location. As agents interact across data, applications and workflows, organizations need infrastructure that operates reliably across data centers, public clouds and edge locations.
Containers are becoming part of that operational foundation. They allow applications, AI models and dependencies to move consistently across environments. That helps explain why 85 percent of global enterprises say AI initiatives are accelerating container adoption.
Agentic AI can generate new business value, but only if it is operationalized with discipline. Without the right controls, the same tools that accelerate productivity can fragment data, duplicate costs and produce unmanageable workflows.
Build before shadow operations scale
Unmanaged workflows surface when the sanctioned path is too slow, complex or disconnected from how the business operates.
When enterprise infrastructure cannot move at the speed of business innovation, unmanaged workflows will keep appearing. Policy restrictions may buy time, but they will not build long-term organizational readiness.
The pressure on leadership will only intensify as AI becomes deeply embedded in daily operations. Enterprises seeking sustainable growth must make the governed path practical enough for the business to adopt. Otherwise, organizations risk losing control of AI workflow by workflow, long before the strategic risk becomes visible.
Faiz Shakir is Vice President and General Manager for Asia Pacific and Japan at Nutanix. The statistics cited in this commentary are drawn from Nutanix-commissioned research.
Editor’s note: This contributed article has been lightly edited for clarity and TNGlobal house style. The substance of the author’s contribution has been preserved.
Share your perspective: TNGlobal welcomes contributed insights and expert commentary from across Asia’s technology and innovation ecosystem. Submit a contribution for editorial consideration, or explore more perspectives in TNGlobal INSIDER and conversations in the TNGlobal Q&A and Interviews archive.
Featured image: Kevin Ku on Unsplash