Would You Get Onboard?

I'm writing this from the airport.

My flight to Charlotte is delayed again while the airline works through a mechanical issue. I'd love to get home sooner, but I have zero interest in watching them rush the process or cut corners to improve an arrival metric. The delay is frustrating, but the priority is clear: make sure the aircraft is safe before it leaves the ground.

Sitting at the gate, I kept thinking about how differently most organizations approach AI.

Over the past year, I've watched companies race to deploy AI across finance, operations, sales, and service. The urgency is understandable. Competitive pressure is real, employees are already experimenting, and boards are asking questions. But urgency often pushes organizations to focus on deployment before they've thought through destination, direction, and risk.

No airline takes off without a flight plan. Before the aircraft leaves the gate, there's clarity on where it's going, who's responsible for what, and what happens if something goes wrong.

Many companies are deploying AI without the business equivalent.

And a real AI strategy isn't a technology roadmap. It's a business roadmap. Where can AI create value? Which opportunities deserve investment? How will we measure success? Without that clarity, you get activity, not outcomes.

But strategy is only half of it. The harder challenge is operational discipline, and it shows up once employees start using AI in everyday work.

An AI forecast shapes a hiring decision. An AI analysis influences pricing. An AI summary lands in a board deck. An AI recommendation affects a supplier negotiation. None of that is inherently a problem. The problem is when no one has defined how those outputs get reviewed, validated, and approved before they drive consequential decisions.

This is where aviation is instructive. Airlines aren't safe because they trust every instrument and every person to be right. They're safe because they assume errors are possible and build disciplined processes to catch them. Checklists exist because people make mistakes. Validation exists because instruments fail. Clear accountability exists because ambiguity gets dangerous when conditions change.

The same applies to AI.

So here's a concrete place to start. Take your last board deck or operating review and ask: which numbers, forecasts, or recommendations in it were generated or shaped by AI? For each one, can you answer three questions: who reviewed it, against what, and who signed off? If you can't, that's not a technology gap. It's a governance gap, and it's worth closing before the next deck.

The companies winning with AI usually aren't the fastest. They're the ones that built the foundations before they accelerated. That work doesn't generate headlines or satisfy the demand for immediate action. But like the mechanic working on my delayed flight, the goal was never just to move quickly.

The goal is to arrive safely.

So as you look at your own AI efforts, one question: if they were an airline, would you feel comfortable getting onboard?

About Root Idea

Root Idea helps CFOs protect the business from AI decision risk. Root Idea works directly alongside finance teams to map AI influence, establishes decision governance controls that hold up to board scrutiny, and delivers training and change management to make governance stick. 

If your organization is scaling AI and governance hasn't kept pace, that's exactly the conversation we're built for. Learn more at rootidea.ai.