The official numbers on AI adoption in Australian small businesses depend on who's asking. According to the Australian Bureau of Statistics, about 12% of all Australian businesses used AI in 2024–25, rising to 22% of medium-sized businesses. According to the National AI Centre's monthly survey of small and medium businesses, 43% reported some level of AI adoption in the December 2025 to February 2026 quarter. Both sources agree on the shape of the story: most use is still drafting, content and analysis; trust, relevance and know-how hold businesses back; and businesses that are already improving how they work are far more likely to use AI. Figures in this article are as at November 2026, from the latest releases we could verify at the source.
Where the figures come from
Three sources are worth knowing. Each asks a different question of a different group, which is why their headline numbers differ.
- ABS Business Characteristics Survey 2024–25. The ABS release came out on 25 June 2026. According to the ABS, the survey covered nearly 7,000 businesses and ran from October 2025 to February 2026. It covers all businesses, including the many very small ones.
- National AI Centre SME AI Pulse. The National AI Centre (part of the federal Department of Industry, Science and Resources) runs a monthly survey through Fifth Quadrant. According to the centre, each wave surveys at least 400 small and medium business owners and decision-makers. Results feed the AI adoption tracker, and the latest quarterly write-up we could find is AI adoption insights: December 2025 to February 2026, first published 7 May 2026.
- NAB SME Business Insights, April 2026. Not government data, but a large and reputable survey. According to NAB's Embracing AI report, it drew on about 670 small and medium businesses surveyed from 10 November to 9 December 2025.
The tracker updates monthly, so check it for anything newer than the figures below.
How many businesses use AI
Across all businesses (ABS). According to the ABS, around 12% of Australian businesses reported using AI in their workplace in 2024–25.
By business size (ABS). According to the ABS media release, around 35% of large businesses and 22% of medium-sized businesses reported using AI, while uptake among small and micro businesses was around 11%. The ABS counts a medium business as one employing 20 to 199 persons, which covers many of the established, growing businesses we work with.
Among small and medium businesses (National AI Centre). According to the National AI Centre, 43% of SMEs reported some level of AI adoption across December 2025 to February 2026, a little below the 45% recorded for September to November 2025. According to the same write-up, adoption reached 44% in February 2026.
By industry. According to the ABS, adoption was highest in information, media and telecommunications (38%), and in professional, scientific and technical services and financial and insurance services (24% each). According to the National AI Centre, fewer than 30% of businesses in construction and agriculture were adopting AI. According to NAB's report, adoption was 35% in manufacturing and 21% in transport and storage, the lowest of the industries it measured.
Why do the numbers differ so much? The ABS counts every business, and most Australian businesses are very small. The National AI Centre and NAB survey small and medium businesses and count "some level" of use, which includes a manager trying a chat assistant. Neither is wrong. They are measuring different things.
What businesses are using AI for
According to the National AI Centre, among SMEs using or planning to use AI, content generation and data analytics lead (54% of adopters each), followed by cybersecurity and threat detection (48%). The centre describes these as productivity uses that are relatively low risk to try.
The same write-up notes that more complex uses, including agentic AI (AI that takes steps in your systems), supply chain improvement and AI-assisted HR, "remain largely untapped". Most businesses are using AI next to their work, for writing and analysis, rather than inside the processes that run the business.
On safeguards, according to the National AI Centre, the most common practice among AI users is checking AI outputs before they affect customers, which about half of current users have in place.
What holds businesses back
According to the National AI Centre, three barriers explain why more than half of SMEs haven't meaningfully adopted AI:
- Trust. According to the centre, around 65% of non-adopting businesses cited distrust in AI decision-making or a strong preference to keep people in control of their business processes.
- Relevance. According to the centre, 54% of non-adopting businesses said AI isn't relevant to their business. The centre reads this as a lack of relatable examples rather than a rejection of AI.
- Not knowing where to start. According to the centre, 19% of SMEs said they don't know how to use AI in their business.
Two findings from the ABS sit underneath these:
- Skills. According to the ABS, 35% of businesses reported a shortage of skills within the business in 2024–25, and of those, 57% said specialist skills or knowledge were required.
- Using data to decide. According to the ABS, only 10% of businesses actively collected or analysed data to make informed decisions in 2024–25, down from 24% in 2021–22.
NAB's report adds an observation rather than a figure: operationally heavy industries such as wholesale, transport and storage, and manufacturing are adopting AI more cautiously, "often constrained by legacy systems, cost pressures, or implementation complexity".
What this means for an established, growing business
The data points the same way as what we see in businesses running on several systems and spreadsheets.
The trust barrier is reasonable, and it has a practical answer. The businesses that hesitate want a person in control. Good AI work keeps one there: AI does the first pass on a repetitive task, a person approves, and nothing reaches your accounts, payroll or customers unchecked. That's how we build it, and it's how the federal Guidance for AI Adoption describes it too.
"Not relevant" often means "not visible yet". For a builder, a wholesaler or a manufacturer, the relevant uses aren't a chatbot on the website. They're reading delivery dockets, supplier invoices, orders and timesheets, and preparing them for a person to approve. Our AI that reads dockets, orders and invoices post shows what that looks like.
Data and systems come before AI. The ABS finding that few businesses use data to decide, and NAB's note about legacy systems, match a common pattern: the information is stuck in separate systems (data silos) and has to be retyped or reconciled before anyone can use it. AI can't analyse what it can't reach, and it can't fix numbers that disagree. Often the first step is connecting systems and cleaning the data so there's one set of numbers, then adding AI to the steps where the input is messy.
Skills are a reason to start small, not a reason to wait. You don't need an AI team. You need one well-chosen task, a fixed trial period and a clear way to judge it. Our guide to running an AI pilot safely sets out the steps.
Where to start
If you're being asked "what are we doing about AI?", here's a sensible order:
- Check your fit. The AI fit check takes a few minutes and tells you whether your first step is AI, automation, integration or data clean-up. If you're unsure about the difference, read AI, automation or integration? Which one your business actually needs.
- Work through a checklist. The AI readiness checklist covers processes, data, systems and people.
- Get an independent view. Our AI readiness assessment looks at how work moves through your business and which tasks are worth trying first. It's delivered through our fixed-price $5,000 Process & Systems Audit.
- Test one task on real work. The back-office AI pilot runs one task with a person approving and a fixed end date.
Our AI consultant for small business page explains how we approach practical AI, from the first conversation to a working workflow.
Frequently asked questions
What percentage of Australian small businesses use AI?
It depends on the source. According to the ABS, around 11% of small and micro businesses used AI in 2024–25. According to the National AI Centre, 43% of small and medium businesses reported some level of adoption in December 2025 to February 2026. The ABS counts all businesses; the National AI Centre surveys SME decision-makers and includes light use.
What is the National AI Centre's AI adoption tracker?
It's a public dashboard from the National AI Centre showing how Australian small and medium businesses use AI, built from a monthly survey run by Fifth Quadrant. It covers adoption, uses, expected outcomes, benefits and challenges, and responsible AI practices.
What are the main barriers to AI adoption for SMEs?
According to the National AI Centre, the biggest are trust (wanting people in control), not seeing the relevance, and not knowing where to start. ABS figures also show skills shortages and limited use of data for decisions.
Is my business behind if we don't use AI yet?
Not necessarily. Many businesses get more from connecting their systems and fixing their data first. Once the numbers are in one place, AI can be added to the tasks where it clearly helps.
When is the next ABS data on business AI use?
The ABS has said its 2026–27 Business Characteristics Survey will include more questions on how businesses adopt and use AI. In the meantime, the National AI Centre tracker updates monthly.
If you'd like to talk through where AI would actually help in your business, and what needs to be in place first, book a free initial chat.