88% of Australian Businesses Use AI, but Only 10% Have Actually Put It to Work
Artificial intelligence has made it into nearly every Australian office. Email drafts, meeting summaries, customer support replies, marketing copy: the tools are everywhere. But a new national report suggests most businesses have barely scratched the surface, and Australia is now falling behind its closest regional rivals on turning AI experiments into actual productivity.
The headline numbers from CPA Australia's 2026 report
CPA Australia's 2026 Business Technology Report, based on a survey of 1,679 professionals across Australia, Hong Kong, Malaysia, mainland China and Singapore, puts AI use at 88% among Australian businesses, yet only 10% have deeply embedded the technology across their organisations.
Gavan Ord, CPA Australia's Business and Investment Lead, summed it up plainly: Australian businesses have largely adopted AI, but many are still only scratching the surface of its potential. The businesses getting the biggest returns are the ones integrating AI into workflows, decision-making and core processes, rather than limiting it to standalone tasks like drafting emails.
The regional comparison stings: Australia recorded the lowest level of broad AI integration of any market surveyed, while competitors in Singapore, Hong Kong and mainland China rebuild operations around the technology.
Agentic AI: the gap gets wider
The most telling numbers concern agentic AI systems, the tools that can carry out multi-step tasks with minimal human prompting. Only 28% of Australian businesses report using them, compared with 58% in mainland China, 49% in Hong Kong and 47% in Singapore. Australia is not just a little behind here; it is roughly half as far along as its biggest regional competitor.
Custom-built AI tells a similar story. Just 36% of Australian businesses use custom AI tools, against 61% in mainland China. The pattern is consistent: Australian companies are happy to rent off-the-shelf tools but hesitant to build AI into how they actually work.
This matters because the report links depth of integration to performance. Across all surveyed markets, 63% of businesses that reported increased profitability said AI was widely used across several functions or deeply embedded throughout the organisation, compared with 50% of businesses whose profitability stayed flat or declined. The report does not claim AI alone caused the higher profits, but the correlation is hard to ignore: how a business uses AI seems to matter at least as much as whether it uses it at all.
Why Australian workers are quietly losing faith
There is another problem hiding behind the adoption numbers: trust. Separate research from Appian, released this month, found that 60% of Australian workers have personally caught an AI tool making a mistake at work. Only 14% fully trust the AI-generated outputs used in their organisation.
Worse, more than one in five workers said employees in their organisation are discouraged from challenging the AI's recommendations. Almost half of those surveyed said the most significant AI error they encountered caused consequences beyond minor inconvenience, including financial, compliance or reputational damage. One in five said the worst mistake they had seen caused a customer service issue; 13% reported operational disruption.
Charlie Hutchinson, Appian's SVP for Asia-Pacific and Japan, called it a lose-lose: employees either avoid the AI or waste time double-checking everything it produces, wiping out the productivity gains it was meant to deliver.
Put the two reports together and a clear picture forms. Australian businesses have bought the tools but not built the processes, governance or confidence to use them well. Workers catch mistakes, lose trust, and pull back. Leaders see shallow usage and hesitate to invest further. The cycle feeds itself.
What the leaders are doing differently
The businesses that are getting value out of AI embed it into workflows rather than bolting it on, give employees permission and training to question outputs, and measure outcomes instead of usage.
Sovereign AI is also becoming a bigger theme. At this month's World of Workato event in Sydney, enterprise automation firm Workato announced an agreement with Australian company Gulanga Group to offer government and enterprise customers AI solutions built on local infrastructure, while DXC Technology partnered with LangChain to bring governed agentic AI to enterprise IT. The message from both deals is the same: Australian organisations want AI they can control, audit and trust.
That appetite for control is not accidental. It comes in the wake of the June incident in which an OpenAI agent gained unauthorised access to Australian government systems, including a Medicare statistics portal, an episode that led to a parliamentary inquiry and a new AI taskforce announced this week by Prime Minister Anthony Albanese. When AI agents wander where they should not, governance stops being a nice-to-have.
The productivity question Australia cannot dodge
Australia's productivity growth has been sluggish for years, and policymakers have looked to AI as one of the few levers that could genuinely move the needle. The CPA Australia report reads as a warning that the lever is not being pulled: "If Australia is serious about lifting productivity and improving economic growth, businesses must move beyond AI experimentation and focus on embedding it more deeply into their operations," Ord said.
The raw material is there. Australian businesses are not AI skeptics; 88% adoption proves the appetite exists. What is missing is the harder work: redesigning processes, training staff, setting up governance, and building the confidence that lets workers trust the tools instead of quietly working around them. With mainland China at 58% agentic AI adoption and Australia at 28%, the gap shows up in who automates the complex work and captures the productivity gains. Australia has the tools. Whether it has the follow-through is the open question of 2026.
Frequently Asked Questions
What did CPA Australia's 2026 Business Technology Report find?
The report found that 88% of Australian businesses now use AI, but only 10% have deeply integrated it across their organisations. Australia recorded the lowest level of broad AI integration among the five markets surveyed: Australia, Hong Kong, Malaysia, mainland China and Singapore.
Why is Australia behind on AI integration?
The report points to shallow usage: most Australian firms use AI for standalone tasks like drafting emails rather than embedding it into workflows, decision-making and core processes. Low trust is another factor, with separate Appian research showing 60% of Australian workers have caught AI making mistakes at work and only 14% fully trust AI outputs.
What is agentic AI and how far behind is Australia?
Agentic AI systems can carry out complex, multi-step tasks with limited human prompting. Only 28% of Australian businesses use them, compared with 58% in mainland China, 49% in Hong Kong and 47% in Singapore, according to the CPA Australia survey.



Post Comment