Home
You’re currently viewing our content for All Regions.
You can switch to a different region:
May 5, 2026
There’s a shift happening with AI in the Australian market. Not a dramatic pivot but a correction.
Call it wave two.
The initial surge was predictable. Curiosity turned into experimentation, and experimentation turned into a rush to “do something with AI.”
But now, that enthusiasm is being met with a harder question: Where’s the value?
According to MiT, 95% of AI use cases are delivering zero business value. These findings have landed hard – particularly with CFOs now asking tougher questions. And rightly so.
Because what we’re seeing isn’t a failure of AI. It’s a failure of intent.
There’s no shortage of AI activity in Australia right now. If anything, there’s too much of it because the real issue is misalignment.
Organisations are pursuing use cases that don’t tie back to measurable business outcomes. They’re solving technical problems, not business problems. Faster code, more content, better automation – these are all valid, but only if they move a core metric.
If your AI investment doesn’t move one of those needles, then what exactly is it doing?
This is where most programs break down. Not at the model layer. Not in infrastructure. But at the very top – in problem definition.
AI is lowering the barrier to entry across almost every discipline. That sounds like a good thing but there’s a catch.
It’s enabling more people to do more things – without necessarily increasing their expertise.
Business analysts are writing code. Marketers are building applications. Engineers are generating content. On the surface, it looks like productivity. But in many cases, it’s actually job creep masquerading as progress.
That gap matters.
Because businesses aren’t just chasing output – they’re chasing quality, efficiency, and scale. And when AI is used outside of core expertise, it often introduces more cost than it removes.
The opportunity isn’t to expand your lane. It’s to optimise within it.
There’s a tendency to look at AI challenges through a technical lens – data readiness, security, governance. Those things matter but they’re not the primary issue.
The primary issue is planning.
We’ve seen this before. Digital transformation promised the same step change. Most programs delivered something but very few delivered shareholder-level impact.
AI is following a similar path.
If 95% of use cases aren’t delivering value, it’s not because the models don’t work. It’s because the use cases were never properly defined, prioritised or measured.
You can’t retrofit value into a project that didn’t start with it.
For example, ask most organisations if their data is ready for AI, and the answer is usually ‘we’re getting there’.
The reality is more blunt.
Most organisations don’t just have a data problem, they have a context problem. The same dataset can mean completely different things depending on who’s using it.
AI doesn’t fix that. If anything, it exposes it.
Without proper context and grounding, AI outputs become inconsistent, unreliable, and ultimately unusable. That’s why so many early deployments fall short – they’re being applied to data environments that were never designed for this level of interpretation.
Despite the noise, there are areas where AI is delivering real value. Unsurprisingly, they’re concentrated in IT.
Recent industry data shows that over half of GenAI use cases are in IT operations, with a significant drop-off into functions like marketing and finance. That’s not because those functions lack opportunity – it’s because IT can more easily quantify value.
These are known levers. But the more interesting work is happening in organisations that are pushing beyond that – particularly in sectors like banking, where AI is being embedded into core processes with clear commercial outcomes.
That’s where the next wave of value will come from.
The organisations that are seeing success in Australia tend to share three traits:
Get these three right, and a business can dramatically improve the odds of being in that top tier.
From a technology standpoint, the AI ecosystem is stabilising. The core components – models, orchestration layers, interfaces – are starting to form a coherent stack. That reduces risk. It gives organisations a more stable foundation to build on.
But it doesn’t solve the fundamental problem of clarity.
Because no matter how mature the platform becomes, it won’t fix a poorly defined use case or a misaligned business objective.
The bottom line is that AI in Australia isn’t failing. But it is being misunderstood. We’re moving out of the phase where doing AI was enough. Now it’s about delivering outcomes.
That requires discipline. It requires focus.
And most importantly, it requires organisations to stop asking what AI can do – and start asking what it should do.
Because in this next phase, the winners won’t be the ones doing the most with AI. They’ll be the ones doing the right things.
Lachlan White is CTO of LAB3. As part of Moxie Top Minds, Lachlan contributed to AI Outlook: Australia 2026 by Moxie Insights. Download the report here.
Inform your opinion with executive guidance, in-depth analysis and business commentary.
