From tools to orchestration: AI's place in the business system of technology across ANZ

A small or mid-sized business in Australia or New Zealand is no longer asking whether to use AI. Most already have, in one form or another, whether that is a marketing draft written in ChatGPT, a bank feed reconciled automatically in Xero, or a customer email answered by a chatbot after hours. The more useful question, and the one most businesses have not yet answered, is whether AI sits inside their business as a connected part of how work actually gets done, or whether it sits alongside everything else as one more disconnected tool. That distinction is where the next layer of Automateley's business system of technology lives, and the evidence from across the region shows both how far adoption has come and how far it still has to go before AI behaves like a genuine part of the system rather than an add-on to it.

Adoption in Australia has moved fast, and the definition of "using AI" is doing a lot of work

The headline numbers on Australian AI adoption vary enormously depending on who is measuring and how, and it is worth being upfront about that rather than picking whichever figure sounds most dramatic. The most rigorous baseline comes from the Australian Bureau of Statistics, whose 2024 to 2025 Business Characteristics Survey found that around 12 percent of Australian businesses reported using AI in the workplace, with adoption accelerating sharply compared with the very low rates recorded in 2021 to 2022 before generative AI tools became widely available, and that almost half of all businesses were finding efficiencies through innovation and new technology more broadly. The federal government's own AI Adoption Tracker, run monthly by the National AI Centre in partnership with Fifth Quadrant, offers a more granular small-business figure, with SME adoption reaching 44 percent in February 2026, its strongest result in several months, as businesses that have already committed to AI tend to expand its use rather than retreat from it. Vendor-commissioned research paints a higher picture again. Intuit's analysis of more than five million QuickBooks businesses alongside a 34,000-business survey found that regular AI use among Australian SMEs rose from 40 percent in July 2024 to 69 percent in January 2026, with the fastest growth in accounting, administration and customer communication.

The gap between these figures is not a contradiction so much as a definitional problem, and it is one worth naming plainly. Surveys that count any use of a free chatbot report far higher numbers than surveys that count sustained, embedded use in daily operations, and a business owner reading the headlines could reasonably conclude that Australia is either barely started or already finished with AI adoption, when the honest answer is somewhere in between and depends entirely on what is being measured.

The real story is not adoption, it is maturity

The more revealing data does not ask whether a business uses AI at all, but how deeply that use has been embedded into how the business actually runs, and here the picture is far more consistent across sources. Deloitte Access Economics, in a report commissioned by Amazon that surveyed more than 1,000 Australian SMBs against a purpose-built AI Maturity Index, found that while roughly two-thirds of SMBs are using AI in some form, only 5 percent are fully enabled to realise its potential, meaning an AI strategy embedded in core processes, staff trained to use it, and a genuinely centralised data system underneath it. The economic weight behind that maturity gap is considerable. The same modelling found that a business moving from basic to intermediate AI use could expect a 45 percent increase in profitability, rising to a 111 percent increase for a business moving from intermediate use to being fully enabled, and that if just one in ten under-adopting SMBs advanced a single rung on that ladder, it would add an estimated $44 billion to Australian GDP annually.

That gap between using a tool and running a system is precisely the distinction the business system of technology is built around, and the barrier data backs this up. The National AI Centre's tracker found that among Australian SMEs not planning to adopt AI in the next twelve months, around 65 percent cited either a distrust in AI decision-making or a strong preference to maintain human control over their business processes as the reason. That is not a technology problem. It is a governance and integration problem, and it points to exactly why AI adoption stalls at the experimentation stage for so many businesses: a standalone tool asks an already stretched owner to supervise it, while a properly orchestrated workflow, embedded in context and governed from the outset, is the thing that actually earns their trust.

New Zealand tells a closely related story, with its own governance gap

New Zealand's SME sector shows the same pattern of fast, patchy adoption running ahead of embedded, governed use. The Ministry of Business, Innovation and Employment commissioned a survey of 500 SMEs which found that awareness of AI tools is very high, at 94 percent, though adoption and confidence vary significantly between businesses of different sizes and sectors. MYOB's Business Monitor survey, run across more than a thousand New Zealand business owners and decision-makers, found that proactive AI use rose from 32 percent to 36 percent year on year, with a sharp divide by size: businesses with one to five employees sat at 30 percent, while those with twenty or more employees reached 64 percent. Xero's own research among Kiwi SMEs found a similar proactive-use figure of 61 percent, with almost half of respondents saying AI could be the most significant opportunity for their business since the rise of the internet, even as most were still learning through self-guided experimentation rather than a structured rollout.

The governance gap is where New Zealand's data is most instructive for a business thinking about how to adopt AI responsibly. A survey of more than 300 businesses by the Employers and Manufacturers Association found that 83 percent were already using AI in some capacity, yet only 13 percent had a formal AI policy in place, with most adoption happening informally as individual staff found tools that made their own work faster rather than through any organisational decision. That is the shadow-AI problem in miniature, and it is a direct illustration of why AI without orchestration is a liability rather than an asset. A business where AI use is scattered across individual staff, each with their own tool and no shared governance, has all of the exposure of AI adoption and almost none of the coherence that makes it safe or valuable at the level of the whole business.

The wider shift: from single AI tools to governed, multi-agent orchestration

The same distinction between a bolted-on tool and an embedded workflow is also where the broader enterprise AI market is heading, and it validates the same principle at a larger scale. Gartner, the most closely watched analyst voice in this space, has forecast that 40 percent of enterprise applications will embed task-specific AI agents by the end of 2026, up from under 5 percent a year earlier, marking a shift from AI as individual productivity assistance toward AI as a connected part of how workflows are executed across a business. The same research describes this as a genuine shift in how enterprise software works, from tools that support individual tasks toward platforms where multiple specialised agents coordinate a whole process, with a human retaining oversight and final authority throughout.

Gartner has also been candid about where this goes wrong, and the warning is directly relevant to any business considering AI adoption without a plan for how it will be governed. The firm has cautioned that more than 40 percent of agentic AI projects could be cancelled by 2027 due to unclear business value, escalating costs or inadequate governance, a finding that echoes almost exactly what the Australian and New Zealand SME data already shows: adoption without structure produces exposure and disappointment, while adoption with clear context, oversight and integration produces the profitability gains the Deloitte and MYOB modelling describe.

Why this belongs inside the business system of technology, not beside it

Read together, the regional data and the industry evidence point to the same conclusion from several directions. Most ANZ small and mid-sized businesses have already tried AI in some form, and a shrinking number remain unaware of it. What separates the businesses seeing a genuine return from those still experimenting is not which tool they picked, but whether that tool was given the context of the specific business it serves and folded into a workflow someone is accountable for, rather than left to individual staff to adopt informally with no policy behind it.

This is why AI orchestration is not a separate offering sitting apart from the rest of the business system of technology, it is the newest expression of the same idea that runs through the whole model. A business's connection, network and security already work best when they are designed and run together rather than bought piecemeal from separate vendors, and the same is true of AI. Built with the context of the specific business it serves, connected to the same systems already carrying that business's infrastructure and security posture, and delivered with a human retaining final oversight, AI orchestration becomes the layer that carries the operational load, and the opportunity is largest wherever a business runs on repeated, high-volume operations that follow a consistent pattern, whatever industry that business happens to sit in. That is the difference between adding an AI tool to a business and giving a business an AI-run layer of its own system of technology, and the evidence from across Australia and New Zealand shows plainly which of the two is actually producing results.

References

1. Australian Bureau of Statistics, Business adoption of Artificial Intelligence accelerates in 2024 to 2025. https://www.abs.gov.au/media-centre/media-releases/business-adoption-artificial-intelligence-accelerates-2024-25
2. National AI Centre (Australian Government), AI adoption insights: December 2025 to February 2026. https://www.ai.gov.au/news-and-insights/blog/ai-adoption-insights-december-2025-february-2026
3. IBS Intelligence, reporting on Intuit AI adoption research across Australia, the US, UK and Canada. https://ibsintelligence.com/ibsi-news/ai-adoption-surges-among-australian-smes/
4. Deloitte Access Economics, The AI edge for small business (commissioned by Amazon). https://www.deloitte.com/au/en/about/press-room/ai-edge-small-business-increased-smb-ai-adoption-can-add-44-billion-australias-economy-251125.html
5. Ministry of Business, Innovation and Employment (New Zealand), AI adoption by New Zealand small and medium-sized businesses. https://www.mbie.govt.nz/business-and-employment/business/support-for-business/research-and-reports-for-small-business/ai-adoption-by-new-zealand-small-and-medium-sized-businesses
6. NZ Herald, reporting on MYOB's 2026 Business Monitor. https://www.nzherald.co.nz/business/companies/ai-adoption-lags-in-small-businesses-myob-research-finds/premium/EXXP62FA3RAIPAQANATLKXZDWU/
7. Xero, From curious to confident: how Kiwi business owners are adopting AI. https://blog.xero.com/data-insights/kiwi-sme-ai-adoption-research/
8. Employers and Manufacturers Association (New Zealand), survey findings reported via Scoop Business. https://business.scoop.co.nz/2026/07/09/ai-adoption-surges-among-smes-but-businesses-risk-falling-behind-on-workforce-readiness-ema-survey-finds/
9. Gartner, Gartner Predicts 40% of Enterprise Apps Will Feature Task-Specific AI Agents by 2026. https://www.gartner.com/en/newsroom/press-releases/2025-08-26-gartner-predicts-40-percent-of-enterprise-apps-will-feature-task-specific-ai-agents-by-2026-up-from-less-than-5-percent-in-2025
10. Forbes, reporting on Gartner's agentic AI project cancellation forecast. https://www.forbes.com/sites/markminevich/2025/12/31/agentic-ai-takes-over-11-shocking-2026-predictions/

This article references publicly available government, vendor and industry research current as of mid-2026. Figures are drawn from the sources listed above and are cited inline throughout. Adoption statistics vary significantly across surveys due to differing definitions and methodologies; where sources conflict, this article favours government and primary vendor research over secondary aggregation.