Where Does AI Ownership Sit? IT, Operations or Somewhere New?
AI now touches every department, yet no department truly owns it. IT manages systems, Operations manages workflows, Strategy sets direction, and HR develops people. But AI sits across all of them.
Every organisation is now wrestling with the same question. AI touches every department, yet no department feels it truly belongs to them. IT manages systems, but AI is no longer just a system. Operations manages processes, but AI now reaches across far more than workflows. Strategy sets direction, but AI requires day to day oversight and governance. HR manages people, but AI transforms how people work. AI is woven through every corner of the organisation, which means ownership often becomes everyone’s responsibility, and as a result becomes no one’s responsibility.
This leadership gap is one of the most significant blockers to successful adoption. It is the gap that allows tactical use to flourish without strategic alignment, and the gap that causes governance to sit on paper rather than in practice. It is also the gap that creates inconsistent adoption, duplicated tools, unmanaged risk and missed value opportunities. Many of the problems organisations experience with AI begin not with technology, but with uncertainty about where ownership should sit. This links directly to the themes in our article on Tactical vs Strategic AI Adoption, and aligns with our article on AI Governance and the Operational Gap.
IT is essential, but it cannot own AI alone
For many organisations, the instinct is to assign AI to the IT department. At a glance it makes sense. IT teams understand systems, integration, procurement and risk. They are already responsible for security and may be involved in vetting tools. Yet AI is not like traditional software. A single IT owned deployment model does not reflect reality when teams across the business are using multiple tools, models and workflows.
AI is also not optimised for the IT mindset. IT is designed to keep systems stable. AI requires experimentation, iteration, change and continuous learning. It requires strategic value alignment, workflow redesign and behavioural adoption, all of which sit outside IT’s natural remit. Most importantly, IT cannot carry the responsibility for training, upskilling, use case development or organisational capability building. These are not technology outcomes. They are enterprise outcomes.
As generative AI expands, IT remains a critical partner, but not the home of ownership.
Operations sees the value, but cannot carry the weight
Operations teams understand where value is created and where efficiency breaks down. They know the workflows that profit depends on. They often see the clearest opportunities for AI to streamline processes, reduce waste and strengthen performance. For that reason, many organisations lean toward operational ownership.
Yet Operations faces its own limitations. They do not typically have the technical expertise to evaluate models, manage data readiness, assess vendor maturity or interpret model risk. Operations also lack the time to train staff consistently or to monitor AI usage across the entire organisation. As the pace of AI evolution accelerates, the burden on Operations becomes too large for one function to sustain.
Much like IT, they are essential partners, but not complete owners.
AI belongs somewhere new
The challenge is not that IT or Operations are incapable. The challenge is that AI does not fit into the organisational structures built in the last twenty years. This shift is more comparable to the introduction of computers, when brand new departments emerged to manage technology that did not fit anywhere yet.
AI is not a new tool or a new channel. It is a new organisational capability. It affects governance, risk, people, customer experience, operational design and strategic value. It touches every link in the value chain. Because of this, AI requires a new organisational home that did not exist before. It requires a place where strategy, governance, capability building and operational execution can be brought together under one clearly accountable leadership model.
This is where we introduce the concept of a central AI function, an AI office or a Chief AI Officer supported by a cross functional council.
The hybrid model: central leadership with distributed execution
Organisations that succeed with AI typically adopt a hybrid ownership structure. They appoint a single owner such as a Chief AI Officer or Head of AI who is responsible for governance, strategy, value alignment and risk posture. This leader is then supported by a council made up of representatives from IT, Operations, HR, Data, Legal and key business units.
This model allows AI to have a clear home for decision making while still benefiting from the context and expertise of every department it touches. The central function sets standards, monitors adoption, tracks value and manages governance, while departments execute use cases within an agreed framework.
This ensures that AI strategy remains consistent, that governance is applied universally and that value can be measured across the entire organisation rather than in isolated pockets. It directly supports the principles we outlined in The Hidden Cost of Not Adopting AI.
The risks of unclear ownership
When ownership is not defined, risk multiplies. Shadow AI grows quickly as employees adopt tools without oversight.
Our article on Shadow AI can be read here.
Tactical adoption begins to dominate because decisions happen at the edges of the organisation rather than in a structured way. Security issues surface as sensitive data finds its way into public models or unmanaged platforms. Departments begin duplicating tools, workflows and subscriptions. No one captures best practice, so value cannot be repeated. Governance becomes a PDF rather than a living process. Leaders struggle to measure ROI or understand how AI is actually being used.
Most importantly, without ownership, AI does not scale. It remains fragmented and fragile.
What clear ownership unlocks
When AI has a defined home, progress accelerates. Organisations are able to build repeatable value rather than one off improvements. ROI becomes measurable and predictable. Teams receive consistent training. Best practices are documented and maintained. Security and risk safeguards become part of daily work. Experiments become designed rather than accidental. Value is aligned to business objectives rather than individual initiative.
Clear ownership turns AI from a collection of tools into an organisational capability.
The Silmaril View
Our perspective is that AI needs central leadership, supported by the organisation but guided by a dedicated function with its own skills, frameworks and governance. It cannot live in IT alone, nor in Operations alone, nor be scattered across departments without clarity. AI represents a profound shift in how organisations work, and with that shift comes the need for a new locus of ownership.
While every organisation is different, many lack the capacity or expertise to establish this new function alone. This is where partnering with specialists becomes essential. A partner can help design the ownership structure, build the road-map, embed governance and support capability building across the business so the function is robust from day one.
The solution is not to decide whether AI belongs to IT or Operations. The solution is to recognise that it belongs somewhere new.