Partner with us for reliable IT support. Contact us now and find out how we can streamline your IT needs!
What governs the pace of AI adoption inside an organisation? What happens if the token price doubles? And what does the board conversation look like when a chatbot's promise becomes a liability? Steve Anderton, Brennan's Head of Customer Solutions, reckons the two-speed AI story most people are telling (vision abundant, value realisation scarce) is only half the picture. The more interesting speeds run inside the fast movers, where four internal clocks decide whether AI adoption sticks or stalls.
Vision on AI is abundant in Australian businesses. Nearly every organisation has one. What’s harder to find is the execution that converts the vision into value.
That gap is one version of the two-speed story. The one I think worth watching runs inside the small group executing well.
“Time to value isn’t decided on go-live day. It’s defined by how well your culture continues to adapt as AI’s capabilities keep evolving.”
Steve Anderton, Head of Customer Solutions, Brennan
PwC's Australian CEO Survey has 45% of Aussie CEOs naming the shortage of internal AI capability as their number one barrier.
Deloitte's Global Human Capital Trends, sampling 9,000 leaders in 89 countries, found 85% of leaders call adaptability critical to their organisation's future. Just 7% say they're leading on it.
And ADAPT's local research puts ~80% of Australian data and analytics leaders on record saying their data isn't ready for the use cases the business is asking for.
The 2026 data confirms the story we keep hearing. Almost everyone agrees AI is critical, but few are executing in a way that realises value.
Inside organisations, adoption is governed by two forces at once. Cultural permission from the top decides whether it can happen at all, with individuals setting the pace once it does.
“Within Brennan, our exponential lift came the moment our MD, Dave Stevens, gave AI his full-throated support. When your MD is the champion, the speed of adoption and change is exponential.”
Steve Anderton, Head of Customer Solutions, Brennan
Without that signal, adoption stays in pockets, no matter how curious your people are. Even with endorsement, the pace still varies wildly. Some race in, even as others hold back.
That's the first internal two-speed: the signal from the top, and the range of speeds that show up underneath it.
The second is governance. While the policy folk are writing the guardrails, the actual users are building in Copilot, Claude and ChatGPT, and aren’t afraid to experiment.
Governance and use need to run in parallel. Halt the tools while you figure out the rules and nothing gets learned about what AI is useful for. Let the tools run without rules and you open yourself to risk.
The third internal two-speed is the one that could reset the maths on all of this: tokenomics.
Tokens are being sold at around one cent on the dollar today, and the frontier labs are already testing whether customers will accept five. So far, the answer coming back is “no thanks”. (Full hat-tip to Cory Doctorow on Jon Stewart’s podcast for the framing.)
“Today’s AI economics look a lot like the subsidy story we lived through with ridesharing and streaming from five years ago.”
Steve Anderton, Head of Customer Solutions, Brennan
If you've built AI into a business process at today's prices and the commercial model shifts, you've got a business-case problem. This is where finance walks into the conversation. Governance came for cloud spend, and we called that FinOps. Tokenomics is the same conversation, one wave later.
The trap in all of this is thinking the challenge is AI adoption. It’s not. Most organisations have a broken innovation process, and AI has just made it visible. POCs wither on the vine because they had no grounding in value. Or they weren't repeatable. Or they took too long. Or they didn't lift beyond the team that had the idea.
What most organisations don't have is the funnel that catches ideas from across the business, presents them for stakeholder review, then delivers a working prototype fast enough to know if the ideas are worth building.
If you can’t spread your institutional intelligence, the same people at the front or in positions of influence will keep doing the same things faster and the gulf widens.
There's another speed challenge showing up across generations. And it was neatly captured by Daniel Hayes, Brennan’s CISO. Recently on the road leading a multi-city roundtable series, he heard Perth participants describe AI as eating the traditional entry points for grads and junior staff in legal and finance.
“My read on what happens next is more optimistic. Graduates entering the workforce today are AI natives. They've grown up with these tools, and their working practices are already built around them.”
Steve Anderton, Head of Customer Solutions, Brennan
For every role AI impacts, new ones will open. Grads and new entrants will continue to be in demand.
The new cohort also bring the ability to look at problems without decades of embedded practice telling them how it’s always been done. A beginner’s mindset – the willingness to ask, “Is there another way? Can we do it better?” – is a valuable trait.
The catch is, experience is a gift, too. Decades of hard-won experience tells you when output of any variety, human or AI, is wrong. Fresh eyes, minus the discernment, is a recipe for fast confusion.
For me, the organisations getting this right pair both: the new cohort bringing the energy to reimagine; the established workers generously sharing their measured, earned wisdom. And when both meet top-down cultural permission, the pace and quality of work lift sharply.
The trick is to replicate what the apprenticeship period teaches, pairing a new entrant with an experienced mentor to ensure institutional IP is absorbed. But they also need the freedom to rethink, reimagine and challenge the status-quo.
Which brings us back to that board conversation. The organisation that thinks it's running fast on AI is often the one that hasn't built the checks that let it run fast responsibly. The 2024 Air Canada case, where a tribunal held the airline liable for what its chatbot promised a customer, is worth a read. You cannot outsource that obligation.
“The two-speed AI adoption problem is really about everything that has to move alongside AI.”
Steve Anderton, Head of Customer Solutions, Brennan
Governance keeping pace with velocity.
Finance keeping pace with "tokenomics”.
The innovation process keeping pace with ideas and being able to deliver on them.
And the people keeping pace with the tools.
If any of them lag, you’ll see it as an AI problem. It isn’t.
I'm living this at home right now. We're in the middle of a major renovation and trying to live in the house at the same time. Multiple things are happening in parallel across the house. The family is camping out on the ground floor before we move upstairs and we’re all hoping it’ll be wrapped by Christmas (the classic building deadline).
But that’s out of my hands. The pace of the build is governed by skills and materials availability, and by what we find behind each wall as we look behind the plasterboard.
AI adoption is similar. Leadership ambition sets the destination. What determines the pace is the skills, the governance, the finance, and the innovation process underneath. You can push, but you can't outrun those constraints. What you can do is build them properly, so what you put on top holds.
For those of us who've spent decades building working practices around certain ways of doing things, it means checking our own ego, being curious and being open to the idea that there is a better way.
Partner with us for reliable IT support. Contact us now and find out how we can streamline your IT needs!