There is good news and bad news when it comes to AI adoption: while most organisations now have access to powerful AI tools, far fewer are getting anything meaningful out of them.
Boston Consulting Group research has found that a large majority of AI transformation efforts fail to meet leaders’ expectations, pointing to organisational culture as the root cause, and not the technology itself.
McKinsey’s research tells a similar story: fewer than a quarter of organisations globally have managed to implement AI at scale, and Australian businesses are following the same pattern.
A familiar story, closer to home
KPMG’s Global AI Pulse survey published in April this year found that Australian organisations are ahead of the rest of the world with regard to responsible AI and governance, but fall behind their global peers when it comes to prioritising AI-driven productivity gains.
In addition, separate research from Insight Enterprises found that only one in five Australian businesses are scaling AI across functions, with most still experimenting or piloting, and fewer than one in ten fully embedding it into their operations.
Altogether, these studies tell a familiar story: Australian leaders are engaged and building the right guardrails, but many are stuck at the pilot stage rather than turning that caution into results.
Where the real gap sits
It’s tempting to treat AI adoption as an IT project: buy the platform, run training, wait for productivity to lift. But the evidence tells us it’s not that simple.
IBM’s most recent Global CEO Study found that while the vast majority of employees now have access to AI tools at work, only a small fraction use them regularly. That gap between access and use might look like a hardware or software problem, but it’s actually a behaviour problem. For AI to be adopted well, employee behaviour needs to be modelled and led from a cultural perspective.
PwC Australia’s CEO research puts a number on this: while 90 per cent of CEOs believe AI adoption is crucial to their business strategy over the next 3-5 years, almost half name internal skills and capability as the single biggest barrier to AI adoption. The Australian HR Institute’s most recent Quarterly Work Outlook points to the same conclusion, flagging a persistent leadership capability gap sitting alongside a growing AI capability gap as a genuine risk to progress.
Few Australian leaders are dismissing AI outright. The more common failure, as consultants working inside local transformation programs describe it, is leaders who want to move fast on AI but feel underprepared to manage the risk of doing so at scale, so they stay comfortable running small, controlled experiments instead.
This means AI often ends up sitting between IT, operations and strategy, with no one clearly accountable for driving it through the business.
Treating AI as a leadership discipline
Leaders need to keep their AI skills up to date in order to leverage the capabilities of AI tools. As the skills grow, the mindset will shift as well. Leaders need to build trust, set clear expectations, and help their people adapt to new ways of working, at a pace few have had to lead before.
Prosci’s change management research shows why this matters so much: organisations rating their leadership support as strong report dramatically smoother AI implementations than those where support is weak, a gap large enough to separate adoption from abandonment.
That’s a leadership development problem before it’s a technology one. The organisations turning AI investment into real results are the ones who assess the needs and gaps, first within the leadership capability, and then the AI software.
