About 12% of Australian employing businesses used AI in 2024–25, according to the Australian Bureau of Statistics, rising to 35% of businesses with 200 or more staff. The money is mostly not going into AI itself: Australian businesses spent $668.3 million on AI R&D in 2023–24, about 2.7% of $24.4 billion in total business R&D.
- Whole economy: 12% of employing businesses used AI in 2024–25 (ABS Business Characteristics Survey, ~7,000 businesses, 87.9% response rate).
- By size: large (200+ staff) 35%; medium (20–199) 22%; small and micro around 11% (ABS).
- SMEs, broader definition: 43% reported “some level of” AI adoption across December 2025 to February 2026 (National AI Centre / Fifth Quadrant SME AI Pulse).
- Medium–large firms: two-thirds of 105 firms in the Reserve Bank’s liaison survey had adopted AI in some form, but nearly 40% described their use as minimal (RBA, June–August 2025).
- Where the spend goes: cyber security and CRM/ERP upgrades led technology spending among RBA liaison firms. AI and machine learning are expected to be “much higher” over the next three years.
- Compliance: the OAIC’s guidance says the Privacy Act applies to all uses of AI involving personal information.
What counts as “AI adoption” in Australian business statistics?
In the ABS survey, a business has adopted AI if it ticked “Artificial intelligence” on a list of technologies it used during the financial year. The ABS defines AI as “a machine-based system that can perceive and learn about its environment and then make relevant predictions, recommendations, content, or decisions”, whether that is software (chatbots, facial recognition) or embedded in devices.
The boundary matters as much as the definition. The ABS says plainly that its question “is not designed to measure intensity or extent of use within the business.” One staff member drafting emails in ChatGPT and a bank running fraud models both count as one “yes”. The survey also covers employing businesses only, and leaves out public administration and education, so sole traders and government agencies are not in the 12%.
An Australian AI adoption figure is a count of businesses that use AI at all, not a measure of how much of the business runs on it.
How it works
Reading an Australian AI adoption figure before you quote it
Find the denominator
Check who was surveyed: all employing businesses (ABS), SME owners (National AI Centre) or medium-large firms (RBA). Quote the one that looks like your organisation.
Check what counts
A ticked box on a technology list and a self-described adoption stage are different bars. Neither measures how much of the business runs on AI.
Separate depth from use
Split minimal use (desktop assistants) from AI embedded in a named process such as forecasting or fraud detection. Most adopters are still minimal.
Match size and industry
Place your business against its own size and industry rows. Then inventory the tools staff already use against the OAIC guidance.
MAKE MORE SALES.
Pay-Per-Result pricing — We scale sales HARD aligned to your interests, better than anyone else.
How many Australian businesses use AI? Three sources, three answers
The three credible Australian measures give 12%, 43% and roughly 67%. They are not contradicting each other. Each has a different denominator and a different bar for what counts as “use”, and that is why the numbers look so far apart.
| Source | Who was asked | Window | What counts as adoption | Figure |
|---|---|---|---|---|
| ABS Characteristics of Australian Business, 2024–25 | ~7,000 employing businesses, all sizes; 87.9% response rate | FY 2024–25 (collected Oct 2025–Feb 2026; released 25 June 2026) | Ticked AI on a list of technologies used | 12% of all businesses; 35% of large |
| National AI Centre SME AI Pulse (Fifth Quadrant) | 400+ SME owners and decision-makers per monthly wave | Dec 2025–Feb 2026 | Self-described “some level of” adoption | 43% for the quarter; 44% in Feb 2026 |
| RBA liaison survey | 105 medium–large firms, guided interviews | June–Aug 2025 | Adopted AI “in some form” | Two-thirds; nearly 40% minimal use |
The decision rule we use, the denominator rule, is simple: quote the figure whose population looks like the organisation you are comparing. Quote the ABS for the national economy or for businesses with 200 or more staff. Quote the SME AI Pulse for owner-run small and medium businesses, and say it is self-reported. Quote the RBA for established medium–large firms, and say the sample is about 100 firms. Quoting 43% as “Australian businesses” overstates adoption more than threefold on the ABS’s own measure (43 ÷ 12 = 3.6).
One detail to watch in the ABS release: its key-statistics line compares the 12% with “1% in 2022–23”, while the body text and the ABS media release compare it with 1% in 2021–22. The media release describes 2021–22 as the year “when generative AI tools such as ChatGPT were not available”, so cite 2021–22 as the baseline.
Want this done for you? We book qualified sales appointments on a Pay-Per-Result basis — you only pay for calls that actually land in your calendar.
AI adoption by business size and industry in Australia
AI use in Australia rises with headcount, and whether a business innovates makes at least as big a difference. The ABS reports 35% adoption among large businesses (up from 9% in 2021–22), 22% among medium businesses (up from 3%) and around 11% among small and micro businesses. Innovation-active businesses used AI at 20%, against 6% for businesses that did not innovate. Among small businesses the gap is 19% against 4%.
| Industry (ABS, 2024–25) | Businesses using AI |
|---|---|
| Information, media and telecommunications | 38% |
| Professional, scientific and technical services | 24% |
| Financial and insurance services | 24% |
| Mining | 18% |
| Health care and social assistance | 17% |
| Arts and recreation services | 16% |
| Electricity, gas, water and waste services | 15% |
| Wholesale trade | 14% |
| Administrative and support services | 12% |
| Rental, hiring and real estate services | 11% |
| Manufacturing; Retail trade; Other services | 9% each |
| Construction | 6% |
| Accommodation and food services | 5% |
| Agriculture, forestry and fishing | 3% |
| Transport, postal and warehousing | 1% |
The fastest climb was in financial and insurance services, which grew 24-fold from 1% in 2021–22 to 24%. The spread across Australian industries runs from 38% in information, media and telecommunications to 1% in transport, postal and warehousing.
What Australian businesses actually use AI for
Neither the ABS nor the RBA publishes a clean percentage split by business function. The nearest thing is the National AI Centre’s SME AI Pulse. Among SMEs using or planning to use AI (not all SMEs), content generation and data analytics each reached 54%, and cybersecurity and threat detection 48%. The Centre calls agentic AI, supply-chain optimisation and AI-assisted HR “largely untapped”.
The RBA’s November 2025 Bulletin fills in the depth among larger firms:
| Depth of use (RBA liaison firms) | What it looks like | Share of firms |
|---|---|---|
| None | No AI adopted in any form | About one-third (the remainder after two-thirds adopted; shares are the RBA’s rounded figures) |
| Minimal | Off-the-shelf assistants such as Microsoft Copilot or ChatGPT for discrete tasks: summarising emails, research | Nearly 40% |
| Moderate | AI assisting specific processes: revenue or demand forecasting, inventory management | Around 30% combined |
| Extensive | AI embedded across several business lines and relied on in critical processes such as fraud detection |
The RBA also found that adoption was “often employee-led rather than employer-led”, that returns on investment were “mixed to date”, and that agentic tools were drawing interest but had low practical adoption. In Australia’s larger firms, the most common form of AI adoption is a chat assistant on a desktop, not AI running a business process.
If we can’t make you money, we don’t deserve yours.
Pay-Per-Result pricing — performance-based alignment.
Where the money is actually going
Most Australian technology money is going into what AI sits on, not into AI. When the RBA asked liaison firms how significant each technology was in their total investment, “the most prominent driver has been addressing cyber risks”. Next came upgrades to internal software such as CRM and ERP platforms, many of which were near end-of-life, and then cloud and data infrastructure. Across the economy, software rose from around 6% of private business investment in 2014/15 to 10.5% in 2024/25.
The one direct, official measure of AI spending is R&D. From the ABS business R&D release (August 2025), worked through:
- AI R&D, 2023–24: $668.3 million; 2021–22: $276.3 million.
- Growth: 668.3 ÷ 276.3 = 2.42, which is the ABS’s “142 per cent” increase.
- Share of total business R&D: 668.3 ÷ 24,400 = 2.7%.
- Against software engineering R&D ($4.9 billion): 668.3 ÷ 4,900 = 0.14, so roughly one AI dollar for every seven software-engineering dollars.
Two cautions. First, R&D is not the same as adoption spending. A company paying for Copilot seats is buying a subscription, and the RBA notes that SaaS is “captured as an operating expense in official data”, not as investment. No official source publishes a national total for what Australian businesses spend on AI subscriptions and implementation. Second, a figure that claims to be that total is modelled, not measured. If you are costing your own programme, see our breakdown of what AI implementation actually costs.
Australian businesses spent $668.3 million on AI R&D in 2023–24, just 2.7% of all business R&D, while cyber security and CRM/ERP upgrades were the most significant items in the technology budgets of the firms the RBA surveyed.
Is my business behind on AI compared with other Australian companies?
Find your size row, then your industry row, then be honest about how deep your use goes. Using the numbers above:
- Size: a business with 200 or more staff that uses no AI is in the 65% of large businesses that the ABS counts as non-users. That is the majority, not the laggard end.
- Industry: a construction business using AI anywhere is already ahead of the 6% industry rate. A software or media business that is not using AI is behind a 38% rate.
- Depth: if your organisation’s “adoption” is Copilot licences, it matches the nearly 40% of RBA firms that call their use minimal. Moving to “moderate” means AI assisting a named process, such as forecasting. In practice that needs an owner and a metric. Operational efficiency metrics for AI covers how to measure that.
Customer-facing processes are where the gap is widest, because they combine the untapped agentic category with the customer-facing governance the National AI Centre says lags. Running one yourself takes a process owner, a CRM integration, call and message compliance under the Spam Act and Do Not Call Register, and weekly QA of transcripts. Budget for at least a part-time owner plus engineering time before anything goes live. This is the category LeadsNow works in: 50,769+ AI-booked sales appointments since 2017, delivered as an AI sales agent service for Australian businesses. Whether to build or hand over depends on your volume and on whether you have that owner.
Where AI adoption meets the Privacy Act
Employee-led adoption, the pattern the RBA found, is the one that creates privacy exposure. Staff paste customer details into whichever tool is open. The OAIC’s guidance on commercially available AI products (published 21 October 2024, updated 17 January 2025) says privacy obligations apply to personal information put into an AI system and to output containing it. As best practice, it recommends that organisations “do not enter personal information, and particularly sensitive information, into publicly available generative AI tools”.
The National AI Centre found the same gap: about half of current SME AI users check outputs before they reach customers, but transparency to customers and ways for customers to raise concerns “lag significantly”. The binding dates and the Acts that apply, including the automated decision-making disclosure that starts on 10 December 2026, are set out in our guide to AI governance in Australia. For most Australian businesses, the first AI compliance task is an inventory of the tools staff already use, not a policy for tools they might buy.
Frequently asked questions
What percentage of Australian businesses use AI in 2026?
The latest official figure is 12% of employing businesses, for the 2024–25 financial year, from the ABS media release of 25 June 2026. Among large businesses it is 35%. The ABS says it will add more AI questions to the 2026–27 survey.
Why do the ABS and National AI Centre figures differ so much?
They measure different populations and different bars. The ABS counts all employing businesses that ticked AI on a technology list, giving 12%. The National AI Centre’s SME AI Pulse asks SME decision-makers to describe their own stage of adoption, giving 43% for December 2025 to February 2026. Neither figure is wrong. They answer different questions.
Which Australian industries use AI the most?
Information, media and telecommunications leads at 38%, followed by professional, scientific and technical services and financial and insurance services at 24% each. Transport, postal and warehousing is lowest at 1%, with agriculture at 3% and construction at 6% (ABS, 2024–25).
Why are Australian SMEs not adopting AI?
Trust is the main reason. Around 65% of non-adopting SMEs cited distrust of AI decision-making or a preference for human control, 54% said AI was not relevant to their business, and 19% of SMEs said they did not know how to use it, according to the National AI Centre’s December 2025 to February 2026 insights.
Does the Privacy Act apply to AI tools my staff already use?
Yes, where personal information is involved. The OAIC’s guidance on commercially available AI products says the Privacy Act applies to all uses of AI involving personal information, covering both inputs and outputs. This is general information, not legal advice.
Pay-Per-Result appointments
See if we’re a fit
We book qualified sales appointments for you and you pay on results, not retainers. Our booking page asks a few quick questions so you find out in two minutes whether that model suits your business.
- 50,769+ appointments booked without cold calling.
- Pay-Per-Result pricing — you pay for booked, qualified calls.
- Pick your own time on our live calendar, no phone tag.