Your AI-Skilled Staff Are Your Quiet Flight Risk

Fifty-nine per cent of Australian workers now use AI at work. The ones who do are twice as likely to be thinking about leaving. If you've spent the past year encouraging AI adoption, that second number should get your attention.

Both figures come from PwC's Global Workforce Hopes and Fears Survey 2026, which asked 1,250 Australian workers and was released on 30 September. AI users are considering a change of employer at 23 per cent, against 11 per cent for non-users.

Put simply, the people you most want to keep are the ones most likely to be looking.

Someone who has worked out how to compress a two-day task into two hours knows they're worth more than they were a year ago. So does every recruiter who finds them on LinkedIn.

The split inside your own business

PwC found 91 per cent of senior executives have used AI at work, compared with 39 per cent of non-managers.

So the AI-fluent person in your team is often not the one you'd expect. It might be the analyst who quietly rebuilt a reporting process, or the coordinator who automated half their admin. They're rarely on a succession plan, because a year ago nobody knew to look.

That's not a technology problem. It's a career design problem.

Why this lands on People and Culture

People and Culture teams are so well placed to help lead AI change in their organisations, and retention is one of the most practical places to start.

If your AI rollout has been run as an IT project, nobody has asked what happens to the people who get good at it. Their work has changed and their market value has changed. Their title and pay probably haven't. PwC found only 7 per cent of non-managers say using AI well is rewarded. And that's often the gap a competitor's offer fills.

For mid-size organisations it's sharper again. You can't always win on salary, so you need to win on growth.

Retention moves that actually work

Find them first. Ask your managers who in their team is genuinely changing how work gets done with AI. Not who has a licence. Who is using it well. You'll likely hear names you didn't expect.

Give them a visible next step. That might be leading a workflow redesign, training their peers, or moving into an adoption role. I've set out what that kind of role involves in what an AI enablement manager actually does, and my guide to every role in an AI team shows where it fits alongside the others.

Recognise the capability, then pay for it. If someone's role has quietly grown into something bigger, say so, and re-scope it properly. Then check what the market pays for that work before your people do it for you. If it's turning into an adoption role, my AI Adoption Lead job description is a practical place to start.

Promote for learning speed, not just tenure. AI fluency and the ability to keep learning now count for as much as years in the seat. If your promotion criteria still reward time served first, your best learners will go somewhere that rewards them sooner.

Build before you buy

In the Bot, Build, Borrow, Buy framework I use with clients, this is a build moment. The people who already know your systems, customers and quirks are the hardest to replace and the cheapest to grow.

Replacing them means going to market against everyone else chasing the same skills. Growing them means a conversation, a plan and a modest budget.

Build their careers now and you keep the capability. Wait, and you'll be hiring it back at a premium.

I'm seeing this firsthand. Top AI talent with real experience deploying AI agents tell me their careers have stalled. Often it's because security settings in larger organisations lock them out of the latest AI and productivity tools, so they can't keep learning. Many of them are now quietly on the market.

If you're working out how to hold on to your AI-capable people, that's a conversation I'm having most weeks. Get in touch and we can talk it through on a quick Teams or phone call.