The next leadership frontier isn’t AI. It’s the rules nobody writes down.
Every organisation we work with is racing to get their AI policy right. Governance frameworks. Ethics guidelines. Approved-tool lists. Training modules on responsible use.
All necessary. None of it is the whole story.
Because while policy is being written in board rooms, culture is being written on the floor — in the daily decisions, the quiet nods, the small compromises that tell people what’s really expected around here.
We call these Unwritten Ground Rules, or UGRs. They’re the informal code that governs behaviour long after the formal policy has been signed off and forgotten. And they don’t disappear just because the problem in front of us is now artificial intelligence instead of safety, or ethics, or engagement. They just find a new subject.
Ask yourself: when someone on your team gets a faster answer from an AI tool but skips the verification step because deadlines don’t wait — what have they learned is really expected? When a frontline employee flags a safety concern and watches nothing change — what have they learned?
Here’s the version I didn’t expect. I was working with a group of leaders recently, and one of them said something the rest of the room recognised but nobody had said out loud: he was holding back his best ideas. Not because he’d run out of them. Because if AI could execute the idea, he worried he’d be handing over the case for his own redundancy.
Think about what that UGR teaches. Not “innovate and we’ll back you.” Something closer to “the smartest move is to make yourself hard to replace, even if it costs the organisation an idea worth having.” Nobody wrote that rule. Nobody would defend it out loud. But it’s exactly what a workplace teaches when AI capability is rising and job security isn’t part of the conversation.
That’s the trap. Leaders are meant to be the ones governing the informal rules for everyone else. But they’re subject to the same unwritten code — and when the code says protect yourself first, the organisation loses the very ideas it most needs, from the people most qualified to have them.
This matters more, not less, as AI capability accelerates. Decisions are being made faster, by more people, with less oversight, using tools most leaders don’t fully understand yet. If the informal rule is keep your best thinking to yourself, that gap between stated ambition and actual behaviour doesn’t stay hidden. It shows up as an innovation pipeline that quietly dries up, at exactly the moment the organisation needs it most.
And leaders are often the last to see it happening. Across the UGRs Stock Take data (where we surface the UGRs) we’ve gathered from organisations worldwide, CEOs typically rate their culture far more positively than their supervisors do on the things that matter most — a gap wide enough that the two groups could be describing different workplaces. That’s not a measurement error. That’s the distance between what leadership believes is happening and what’s actually being lived one level down. If that gap exists on culture broadly, there’s no reason to think it closes the moment the subject turns to AI and job security.
So here’s the frontier question, and it has nothing to do with which AI tools to approve: what are the UGRs in your organisation right now around AI — the ones everyone’s quietly operating by, that nobody’s said out loud?
I’d genuinely like to know what you’re seeing. Drop it in the comments — even if it’s uncomfortable. Especially if it’s uncomfortable.
Back to all insights ←