Who Should Own AI Strategy and Governance in a Mid-Sized Company?
AI belongs to named people, each holding a specific piece of the work, with the executive team meeting on a fixed cadence to make the decisions. AI Transformation Framework (AITF) asks every client for two named roles before the governance sessions start, a program owner and a process owner, then assigns an owner to each of the Four Pillars of AI Governance as the policy gets built. What makes ownership hold is a person's name sitting beside each piece of work and a standing meeting where that person reports.
The question most leadership teams ask first is which title should carry it. That question has no clean answer, because the work splits across four or five people regardless of where the org chart puts it.
Why does ownership stay unclear even when leadership feels confident?
Because confidence and allocation move at different speeds, and the gap between them is where accountability disappears.
Deloitte's Q2 2026 CFO Signals Survey, published in July 2026 and drawn from 200 North American CFOs at companies with at least $1 billion in revenue, found that 96 percent of CFOs express confidence in their company's AI governance framework. In the same survey, 51 percent name a lack of governance authority as an obstacle and 43 percent cite insufficient visibility into which AI tools are in use. Only 19 percent identify themselves as carrying the greatest responsibility for AI governance.
Morgan Stanley's August 2026 report on AI governance, based on a survey of 200 executives involved in AI governance at global companies above $100 million in revenue, found that 90 percent have defined or are defining where responsibility for AI governance sits, while just under half say that responsibility is fully allocated today.
ISACA's March 2026 research, drawn from 681 digital trust professionals in Europe, puts the sharpest number on it: 20 percent do not know who would be ultimately accountable if an AI system caused harm, and 38 percent identify the board or an executive.
Those samples all sit at or above the top of the mid-market. The pattern holds at smaller companies with fewer formal risk functions to fall back on.
Do we need an AI steering committee, and who should sit on it?
A committee holds when every seat carries a named piece of work and the group meets on a set cadence. Both of those come from naming owners first, which is where AITF starts.
The roles AITF asks a client to fill:
A program owner, named before the governance sessions begin. This person carries the program through the year and keeps the AI Impact Dashboard current, and the dashboard stays with them at the end.
A process owner, named at the same time. Two roles, two names.
An owner for each of the Four Pillars of AI Governance, confirmed in the week after the governance workshop. AITF assigns the Amnesty and Transparency pillar to the CEO and the communications function, because the memo that opens honest disclosure carries weight from leadership that it loses anywhere else.
An owner for every barrier the readiness work surfaces, agreed by the end of month one.
That set of names is the working group. It combines the authority to make a decision stick, the technical knowledge to judge a tool, the risk discipline to catch exposure, and the communications reach to bring staff along. Whether a company calls it a steering committee matters less than whether each of those names exists and each one reports monthly.
Who should make the AI platform decision?
The executive team, in the room, with IT informing the choice.
AITF tells a client ahead of the governance session that they will select their organization-wide AI platform during it, then holds the decision itself for the live conversation with a frame and selection criteria attached. Settling it earlier tends to mean IT picks a platform before the executives have heard why the choice shapes everything downstream. Every tool permission rule, every training session, and every initiative in the roadmap is built on that one selection.
The Governance gate at the end of month three requires three things together: a named organization-wide AI tool the client has chosen, a licensed business-tier tenant on it, and a live process for answering new tool requests. All three, or the gate stays open.
Morgan Stanley's August 2026 survey found that 71 percent of those companies now require employees to use company-approved AI tools only. A rule like that holds when someone owns the approved list and someone answers the requests to add to it.
What does ownership look like month to month?
It looks like a round table that opens every monthly checkpoint, with every person speaking by name in a set order: one thing they tried, one thing that worked, and one thing that fell short. A failure is a valid and welcome report, which is what keeps the reporting honest.
The standing items that follow are the three project-managed initiatives and their progress, the decisions made with who made them and when, action items with an owner, a due date and success criteria, and the open risks. The rule AITF runs these meetings on is to bring decisions. Anything short of a decision becomes an action item and the meeting moves on.
Twelve of these sit on calendars from the start of the engagement, which is the part most companies skip. Booking them one at a time is how a twelve-month program quietly becomes a nine-month one.
Where does ownership sit across the twelve months?
Ownership is the first gate and the last one.
At kickoff, the Alignment gate closes when the executive team agrees on urgency and commits to the program. At the end of month one, the Readiness gate closes when readiness is mapped, the top barriers are named, and an owner is agreed for each. At the end of month five, the Roadmap gate closes when a prioritized initiative list has been produced and owned by the client. At the end of month twelve, the Sustain gate closes when the client can keep running without heroics, with the dashboard, the governance framework and the roadmap all staying with them.
AITF runs as a twelve-month coached engagement across two phases and seven gates, built so the leadership team owns the work and runs it independently at the end. The program structure and the full deliverable list sit on the AITF site, and the FAQ page answers what leadership teams ask before committing.
Frequently asked questions
-
The work splits across several people regardless of title, so AITF assigns named owners to specific pieces: a program owner and a process owner for the engagement, an owner for each of the Four Pillars of AI Governance, and an owner for every barrier the readiness assessment surfaces
-
A committee works when each seat carries a named piece of work and the group meets on a set cadence. The seats that matter are the ones holding the four governance pillars, the program itself, and the barriers found during readiness, with the executive team meeting monthly for twelve months.
-
Technical judgment belongs in the room and the platform decision belongs to the executive team. Selecting a platform before leadership has heard why the choice shapes training, permissions and the roadmap tends to produce a tool that IT owns and the business works around.
Twenty minutes with an AITF coach will tell you whether your company is at the stage where this program pays off.
