AI Maturity in Australian Businesses, 2026
In May 2026 we asked 60 senior leaders across 50 Australian businesses to place their organisation on Adaca's Crawl, Walk, Run model of AI maturity. This page is the permanent record of what they told us. The distribution is stark: 45 of the 50 businesses placed themselves at Crawl, 4 at Walk, and 1 at Run, and not one had an AI agent running in production.
How the Data Was Collected
The figures come from Adaca's executive roundtable series: three sessions run in May 2026, one in Sydney, one in Melbourne, and a national virtual room for leaders who could not attend in person. Around a fifth of attendees came from financial services, with the rest spread across other sectors. At each session, attendees placed their own organisation on the Crawl, Walk, Run model by a show of hands. The figures are self-reported. We deliberately did not audit the businesses to verify their placements: the dataset records how Australian leaders see their own AI maturity, and that perception is itself the finding.
The roundtable series continues through 2026. This page records the May 2026 round; later rounds will be published as they are run.
The Distribution
| Stage | What it means | Businesses (of 50) | Share |
|---|---|---|---|
| Crawl (Foundations) | Baseline understanding of LLMs across the business, plus a commercial agreement with a provider | 45 | 90% |
| Walk (Governance) | AI governance in place: risk management and visibility of shadow AI | 4 | 8% |
| Run (Building) | Agents deployed into the business | 1 | 2% |
The single business at Run had built an agent that was not yet in production. It sat with the CTO rather than serving the business, which means the count of AI agents in production across all 50 businesses was zero.
Baseline LLM literacy + a provider agreement
AI governance, risk and shadow-AI monitoring
Agents deployed into the business
What Sits Behind the Numbers
The most technical people in these organisations are moving fastest. CTOs and CIOs told us they are using tools like Claude Code to ship systems in days that once demanded a full quarter, while accounts, operations and the rest of the business sit on a single Copilot licence, and sometimes on nothing at all. The engineering function is effectively at Run for its own work while the operational core has not moved.
Many of the same people who have no sanctioned way to use AI at work use ChatGPT and Claude confidently in their personal lives. The result is a widening internal gap and a steadily growing pile of shadow AI, as staff reach for personal tools to fill the space their organisation has not provided for.
The companies stuck at Crawl were not short on ambition. They were short on the foundations: a clear acceptable-use policy, a commercial agreement with a model provider, basic AI literacy beyond the technical team, and any reliable way of seeing where AI is already in use. A handful had a dedicated AI policy; many more were leaning on existing data policies and hoping they would stretch.
Common Questions
How many Australian businesses have AI agents in production?
Zero of the 50 businesses at Adaca's May 2026 roundtables had an AI agent running in production. One had built an agent, but it was still in testing and sat with the CTO rather than serving the business.
How was the AI maturity data collected?
60 senior leaders across 50 Australian businesses self-assessed against Adaca's Crawl, Walk, Run model at executive roundtables in Sydney, Melbourne and a national virtual session in May 2026. The figures are self-reported and deliberately unaudited.
What is the Crawl, Walk, Run model?
Adaca's three-stage AI maturity model. Crawl is foundations: baseline LLM literacy across the business and a commercial agreement with a provider. Walk is governance: risk management and visibility of shadow AI. Run is building: agents deployed into the business.
Where do most Australian businesses sit on AI maturity?
45 of 50 placed themselves at Crawl, 4 at Walk and 1 at Run. Nine in ten are at the beginning of the journey by their own assessment.
Citing This Data
Cite as: Adaca, AI Maturity in Australian Businesses (May 2026 roundtable series), https://www.adaca.com/insights/ai-maturity-2026/.
The analysis was first published in Nobody in the room had an AI agent running in production.
If you are working out where your own organisation sits, the AI Journey assessment produces a stage placement in about 15 minutes.