One of the first questions that comes up when a growing business decides to get serious about AI is: who’s supposed to be running this? It sounds like a technology initiative, so the instinct is to hand it to whoever handles technology. But in most businesses in the $5M to $50M range, the IT function is either a single person who is already stretched across every system the company runs, or an outsourced relationship that handles maintenance and breaks-fixes and not much else.

Handing an AI initiative to that function usually means the initiative doesn’t move. Not because the person isn’t capable, but because the organizational conditions aren’t right. AI adoption isn’t primarily a technical problem. It’s a business and behavioral problem that requires someone who is motivated to solve it, connected to the day-to-day work the AI is supposed to improve, and willing to push through the friction that any meaningful change creates.

The Person You’re Looking for Is Already in Your Building

Henry Hays has a consistent observation from his work across industries: the right person to own AI inside a company is almost never the person you’d pick based on job title alone. It’s the person who is already curious about AI on their own time.

You probably know who this is. They’re the one who brings up AI in team meetings because they’ve been thinking about it outside of work. They’ve already set up personal accounts with AI tools and figured out how to use them for their own projects. They ask questions about AI capabilities that go a level deeper than what they’d need to know just to do their current job. They’re not waiting to be told this is their responsibility. They’re interested in it because they’re genuinely interested in it.

That person might be in marketing. They might be in operations. They might be in sales, or finance, or customer service. They almost certainly aren’t the person you’d identify if you were searching by job title. And that’s fine, because the job title isn’t what qualifies someone to drive an AI initiative. The motivation and the genuine interest are.

When you find that person and give them the organizational support, resources, and permission to go deeper, you end up with an internal owner who is intrinsically motivated to make it work. That motivation is the variable that determines whether an AI initiative builds momentum or stalls after the first few months.

The Jobs That Don’t Have Names Yet

Hays uses an analogy that puts this in useful historical context. When the automobile arrived in 1908, nobody had ever heard of a car wash. But people were already washing their cars, using the same behaviors that would eventually become the car wash industry. The behaviors existed before the job title did.

The same thing is happening with AI right now. Inside most companies, people are already doing work that will eventually be recognized as formal AI roles: managing AI tools, evaluating outputs, training internal systems, identifying new use cases, governing usage across teams. Those behaviors exist under current job titles, invisible because nobody has labeled them yet.

The person in your company who is already doing these things, experimenting with AI tools, thinking about where they could apply, troubleshooting when something doesn’t work, is doing the early version of a job that will be formalized over the next several years. Finding and investing in that person now gives the organization a head start that compounds over time.

What Internal AI Ownership Actually Requires

Identifying the right person is the first step. The second is giving them what they need to actually move the initiative forward.

That starts with time. AI adoption doesn’t happen alongside a full workload without something giving way. If the internal AI owner is expected to drive this initiative in addition to everything else on their plate, the initiative will be the thing that gets deprioritized when things get busy, which is always. A realistic allocation of time, even if it’s just a few hours a week to start, signals that this is real and not just an additional assignment.

It also requires budget. Not necessarily a large budget, but enough to access the tools, the training, and occasionally the outside expertise needed to move forward with confidence. An internal owner who has identified a promising use case but can’t get approval to spend $200 on a tool is an internal owner who is going to lose momentum fast.

And it requires organizational air cover. The internal AI owner will encounter friction. People who are skeptical. Processes that were built without AI in mind and resist being changed. Stakeholders who want to see results before they’re willing to support the initiative. That friction is normal, and the person driving the initiative needs to know that leadership is behind them when they run into it.

Governance Without a Technical Team

One of the questions lean teams ask most often is how to govern AI responsibly when there isn’t a dedicated person whose job is to think about risk, compliance, and data security.

The answer isn’t to recreate a technical governance function from scratch. It’s to build governance into the usage policy and the workflows from the beginning, so that responsible use is the default rather than something that requires active oversight to enforce.

A clear usage policy, one that specifies which tools are approved, what data can be shared with those tools, and what kinds of outputs require human review before they’re used, handles most of the governance questions that come up in a growing business. It doesn’t require a compliance officer to enforce. It requires clarity, and then it requires leadership to model it.

The internal AI owner can also serve a governance function by staying close to how the tools are being used across the team and flagging anything that looks outside the boundaries of what’s been agreed on. This isn’t a formal audit process. It’s a natural extension of ownership: the person who knows the tools best is also the person best positioned to notice when something is being used in a way that doesn’t fit the policy.

Building Capability That Compounds

The most important thing about getting internal AI ownership right is that it creates organizational capability that compounds over time. Every use case the internal owner figures out, every workflow they improve, every team member they bring along on the learning curve, adds to a base of institutional knowledge that makes the next initiative faster and the one after that faster still.

Companies that rely entirely on outside vendors or consultants for their AI initiatives don’t build that capability. The knowledge lives outside the organization, and when the engagement ends, it leaves with the vendor. Companies that develop genuine internal ownership build something that stays.

The businesses that will have the clearest AI advantage in three to five years are largely the ones that identified their internal AI owner early, resourced them appropriately, and gave them the organizational backing to build something real. That person probably already works for you. The question is whether you’ve found them yet.

Listen to the full podcast episode here.

Related Resources

• Your Employees Are Already Using AI at Work — Here’s What That Means for Your Business: Before the internal AI owner can move forward effectively, the organization needs a shared policy for how AI is used. This post covers what that policy needs to include and why it matters.
• Why AI Adoption Stalls at the Leadership Level (And What to Do About It): Internal ownership works best when leadership is also in the game. This post covers what it looks like when leaders model AI adoption and why it matters for the rest of the organization.
• The Boring, Mundane, and Tedious: Where AI Actually Delivers for Growing Businesses: Once ownership is established, the next question is where to start. This post identifies the workflows where AI delivers the fastest, clearest returns for growing businesses.
• The AI Execution Problem: Why Most Growing Businesses Can’t Get Past Experimenting: Internal ownership is one piece of a larger execution challenge. This post covers the full picture of what it takes to move from AI experimentation to real results.