More than half of Australian workers use AI. Only one in ten say it’s actually built into the way their business runs.

That’s the finding from new Appian research released this week. And it’s not a small gap. It’s the difference between using a calculator and changing how you do the maths.

The report surveyed employees across Australian organisations and found that while AI adoption is widespread, most of it is happening at the edges, customer service (20%), marketing (17%), not in the operational core. Finance sits at 8%. HR at 6%.

Nearly one in three employees said their business still relies on standalone AI tools rather than integrating AI into their actual workflows. Which means they’ve added AI to the toolkit but not changed how the work gets done.

Australia is behind the global average, and we already knew it

The global benchmark from a Harvard Business Review Analytic Services study (also commissioned by Appian) puts workflow-integrated AI at 18% globally. Australia sits at 10%. That’s not a rounding error, it’s a structural gap.

We’ve seen the signals here before. Federal government data identified exactly why Australian small businesses aren’t adopting AI, guidance, skills, and integration support are all missing. The Appian research puts numbers to what that looks like inside actual organisations.

The experiment phase is over for some, and just beginning for others

46% of Australian employees describe their organisation’s AI expectations as optimistic or overly optimistic. Only 23% report significant improvements. 11% say they’ve seen no measurable impact at all.

That gap between expectation and outcome has a name: the AI productivity trap. You buy the tools. People use them. Nothing much changes. Because the tools aren’t connected to the processes, the actual decisions, the actual workflows, the things that determine how a business performs.

Luke Thomas, Appian’s VP for Asia Pacific and Japan, put it directly: “Process is how organisations structure work, make decisions and serve customers. When AI isn’t connected to that, it lacks the context, control and visibility needed to deliver meaningful outcomes.”

That’s enterprise language. But it applies to a three-person bookkeeping firm the same as it does to a 500-seat contact centre.

The top barrier isn’t cost. It’s skills and integration.

When employees were asked what’s stopping deeper AI adoption, skills and talent gaps came first (27%). Integration with existing systems was second (20%). Governance concerns followed at 14%.

This matches what we hear from small business owners. The problem isn’t access to tools, there are more AI tools than anyone can evaluate. The problem is knowing which ones connect to your actual work, and having someone help you wire it in.

If you’re using ChatGPT to draft emails and haven’t touched how you invoice, onboard staff, or handle customer complaints, you’re in the 90%. That’s not a dig, it’s where most businesses are. But the companies pulling ahead aren’t the ones with the best AI subscriptions. They’re the ones that picked a single process and fixed it properly.

What “embedded” actually looks like for a small business

You don’t need a process automation platform to move into the 10%. Here’s what it actually looks like at SMB scale:

Your invoice follow-up runs automatically through Xero, you don’t chase late payments manually anymore. Here’s how to set that up in 15 minutes.
New staff get onboarded through a checklist that AI helped you write and now lives in your comms tool, not in someone’s head. The AI Upgrade series covers exactly this.
Customer complaints get a consistent first response instead of waiting for whoever happens to see the email. Not getting results yet? This explains why.

None of those require budget. They require choosing one process and finishing the job.

The encouraging part

The research also shows where organisations plan to go next. The top priority for the next 12 to 24 months is integrating AI into core systems and processes (32%), ahead of expanding existing use cases (19%) or continuing to experiment (17%).

That’s the right order of operations. Experiment, pick what works, then wire it in.

The businesses getting there first, even small ones, will look back on this period as when the gap opened. Between the ones who dabbled and the ones who changed how they work.

Ten percent is a low bar. But it’s also a real opportunity.

📬 Get the SmallBizAI.au Weekly

Practical AI news for Australian small businesses — every week, free.

Source: Appian Australian AI Adoption Research, June 2026. Harvard Business Review Analytic Services study also referenced (commissioned by Appian).

📖 More deep dives: News Deep Dives

📊 Compare AI tools side by side | 💼 Free resources & AI prompt packs

📬 The SmallBizAI Brief

One practical AI tip for Australian small business. Every Tuesday. Free.

Join business owners getting smarter about AI every week.

Sources