On a Section fireside chat this week, our guest Eric Vaughan said something that made me a little nervous: The Head of AI role doesn’t need to exist.

That’s my role (gulp).

Eric isn’t some pundit - he’s the CEO of IgniteTech, a nine-figure enterprise software company that turned over 80% of its workforce in 2023 for refusing to adopt AI. So when he said this, I listened.

Eric hired a Head of AI in 2024. The person got recruited away, and Eric decided not to backfill the role. One, it was creating a centralization problem. Things were getting funneled to one person that should have been distributed. And two, the role was sending the wrong signal. As he put it on the event: “It made it sound like AI was only one person’s job, and that’s not right.”

He’s not entirely wrong here. Sometimes a company hires a Head of AI, everyone else exhales and assumes it’s handled, and AI adoption becomes one person’s problem instead of a company-wide operating shift. The role becomes a bottleneck instead of an accelerant.

But Eric isn’t operating without a Head of AI. He is the Head of AI. He just doesn’t use the title.

The CEO as Head of AI

IgniteTech is a company where the CEO personally interviewed every new hire to screen for AI DNA. He mandated AI Mondays, an entire day per week where no one was allowed to do non-AI work. 

He ran self-assessments, tracked tool usage across every team, and told people directly: if you’re in the basement on AI adoption, you have a quarter to move up or you’re out. He went into the Perplexity backend and removed every spending limit he could find. He’s in the tools every day, arguing with the AI about its standards (his words, not mine).

That’s not a company without a Head of AI. That’s a company where the CEO absorbed the role entirely because he has unusual energy for it, a technical background, and a company he rebuilt from scratch around AI.

For the first 30 days of an AI transformation, I actually think Eric’s model is right. The CEO should be the one setting the tone, making it clear this isn’t optional, and demonstrating personal commitment. That signal has to come from the top, and Eric is a perfect example of how powerful it is when it does.

But beyond that first month, the CEO as Head of AI breaks down fast.

Running a company is already a full-time job. Tracking tool usage across every team, evaluating every hire for AI readiness, personally managing tool access and spending limits, setting adoption standards and holding people accountable - that’s another full-time job. Eric has the time and inclination to do both. Most CEOs don’t, and honestly shouldn’t. Their job is to run the business. They need someone else whose entire focus is making the organization AI-powered.

What the role actually looks like

The Head of AI role works when the CEO has empowered the person to be relentless about transformation, with the authority to make uncomfortable decisions and the credibility to push back on people who aren’t moving.

In practice, that means four things:

Own the metric. You need someone to own, and be accountable for, making the company AI-powered. Companies measure this in different ways - DAU, AI fitness, etc. - but however you measure it, it can’t sit with a bunch of department leads or it will be abandoned. 

Set the standard and hold everyone to it. The Head of AI defines what “good” looks like for the organization and makes it uncomfortable to fall short. At Section, we’re holding employees to a proficiency bar this year - anyone who doesn’t meet it will be performing below expectations. 

Make sure people have what they need and then remove every excuse. The Head of AI’s job is to make it impossible for someone to say “I would use AI, but I don’t have access / time / training / permission.” This means ample training, time, tools, and resources to facilitate their use of AI. 

Identify what should be built and make sure it reaches the whole organization. This is the responsibility that most CEOs won’t do themselves, and it’s the one that makes the standalone role necessary. At Section, people are constantly building useful AI workflows that live on their laptops. The Head of AI figures out which of those should become shared tools, coordinates the people building in parallel so they’re not duplicating effort, and contains the messiness that comes with everyone building at once.

End game

Eric’s instinct is right about one thing. The endgame isn’t a permanent Head of AI. It’s an organization where AI transformation isn’t one person’s crusade anymore because it’s how everyone operates.

But right now, most companies are nowhere near that. Eric is able to do this because he has a rare alchemy of time, energy, domain knowledge on AI, and vision for transformation. Almost no other CEO has that luxury, especially at larger organizations. For everyone else, there’s the Head of AI.

See you next week,

Michael

Your fellow Head of AI

P.S. If you weren’t able to attend our fireside chat with Eric, I highly recommend watching the recording. He was one of our best guests this year, and the bar is high.