The most valuable thing our Executive gave the AI program wasn’t an approval. It was a posture.
I remember preparing for those early Executive conversations, expecting to be asked for the usual artefacts: the strategy, the business case, the risk assessment, the decision to endorse. Some of that came, in time. But the moment that shaped everything was when the Executive effectively settled a different question. Not “do we approve this program”, but “how do we intend to engage with this technology?”
The posture they landed on had three parts. We will engage with AI rather than wait it out. We will learn before we scale, and we expect management to do the same. And we expect honest reporting of what isn’t working, not just what is.
A decision would have been weaker, and it’s worth spelling out why. Decisions are point-in-time. They freeze understanding at the moment of approval, and AI refuses to hold still. A program “approved” in one quarter is operating in a different technological reality two quarters later. When the world shifts, a decision becomes something to defend or revisit. A posture just keeps applying.
That posture did real work at management level. When things moved fast, I didn’t need to return for re-approval; the direction was clear and durable. When experiments failed, and some did, reporting them honestly was the expectation, not a confession. It shaped culture beneath the Executive too: learning before scaling became language people used unprompted.
There’s a lesson here for directors and executives. Your instinct with new technology will be to ask for the plan and approve it, because that’s the familiar governance muscle. With AI, consider first agreeing how the organisation will hold the technology: the pace, the honesty, the appetite for learning. Get the posture right and dozens of downstream decisions get easier, or become unnecessary.
Ask your Executive what its AI posture is. If the answer is a list of approved projects, the real conversation hasn’t happened yet.