How AI is Being Adopted by Businesses in Australia

Australian per-capita Claude usage already outpaces the United States.

Not in five years. Now. Alongside South Korea and Singapore, Australia is one of the highest per-capita Claude markets in the world. And yet, if you spend your days working with B2B companies across ANZ, the AI adoption story looks different to what's being reported out of San Francisco.

The US narrative says enterprise AI adoption is accelerating everywhere, uniformly, and fast. The numbers back that up at a surface level: Australia's AI market is valued at $7.25 billion in 2025, 49% of Australians report using generative AI in the past 12 months, and Anthropic, OpenAI, and Google have all announced local hiring within months of each other. A big concentration of frontier AI platforms entering one market at the same time.

But the reality on the ground in ANZ B2B is more nuanced than the headline figures suggest.

Over the last 12 months, I've worked closely with around 20 B2B companies across Australia and New Zealand, from 30-person scale-ups to enterprise organisations. Financial services, property technology, professional services, SaaS. Across all of them, I've watched AI adoption play out, and the patterns emerging tell a story. One that any AI platform trying to build a serious presence here needs to understand.

This is what I'm actually seeing.

The Gradient of Adoption

AI adoption across ANZ B2B isn't random. It follows a gradient, and the companies that try to skip steps are the ones that stall.

Marketing and content is where everyone starts. Low risk, high visibility, easy ROI. Draft an email sequence, repurpose a webinar into LinkedIn posts, generate a first draft of a case study. The output is tangible, the downside minimal, and the person doing it usually sits in marketing or is the founder wearing the marketing hat.

Around 65% of the companies I work with started their AI journey here. A few stayed here. Most moved on.

Sales and revenue operations is the second wave. Once content works, ambition grows. Teams start asking whether AI can help with lead scoring, call summaries, pipeline forecasting, outreach personalisation. This is where it gets more interesting and more complicated. AI is now touching revenue, and that changes the decision-making dynamic. Approvals take longer. The questions get harder.

The typical timeline from "AI for content" to "AI for sales ops" is three to nine months. The ones moving faster usually have a RevOps function, even a small one, that can own the implementation. The ones moving slower are still figuring out data hygiene.

Customer lifecycle and service is still early. Using AI for retention prediction, churn signals, automated lifecycle communications, or support triaging is theoretically compelling. In practice, most mid-market companies aren't there yet. The blocker isn't desire. It's data readiness. CRM data is messy, customer segments aren't clean, and the integration complexity required is beyond what lean teams can absorb while running the business.

One SaaS company tried to implement AI-driven support routing and spent three months on data cleanup before the actual implementation could begin.

Operations and back office AI is barely started. Finance automation, AI in procurement, HR applications. Huge theoretical potential. Minimal actual adoption across the ANZ market. Compliance concerns, legacy systems, and integration complexity are the real blockers. For most companies, it's not on the roadmap for the next 12 months.

The Deloitte State of AI 2026 report backs this up: only 28% of Australian companies have moved 40% or more of their AI pilots into production. And according to CSIRO-cited research, 80% of AI projects fail to progress beyond pilot stage altogether, double the failure rate of conventional IT projects. The gap isn't enthusiasm. Everyone's enthusiastic. The gap is in implementation capacity and sequencing.

The companies succeeding are building sequentially. 

How Australia Differs From the US

The US AI adoption narrative is built on a fundamentally different business landscape. Applying it to ANZ B2B is like using a map of Manhattan to navigate Sydney. The roads look similar until you actually try to drive.

Team size and structure changes everything. Even large ANZ organisations run leaner than their US counterparts. At the mid-market end, there's no dedicated AI team, no data science department, no transformation office. At the enterprise end, there often is, but it's smaller than you'd expect and more stretched. The person leading AI adoption is frequently doing it alongside a full existing remit, whether that's the head of marketing, the CTO, or a RevOps lead with ten other priorities.

When I've watched AI implementations succeed in these businesses, they've succeeded because one person owned it and had the headspace to see it through. When they've failed, it's usually because the project was everyone's responsibility and therefore no one's.

The partner ecosystem is thin. In the US, there are hundreds of AI implementation partners, consultancies, and system integrators. In ANZ, the ecosystem is sparse. Companies adopting AI often have no local expert to help them go beyond the obvious use cases. They're figuring it out themselves, which works until it doesn't.

This is a genuine unmet need. The demand for implementation support is real and growing. The supply of people who can deliver it well is not keeping pace. Anthropic's own CCO Paul Smith has said publicly that enterprise deployment requires "both large global systems integrators and niche consultancies" trained to implement and build on the platform. That ecosystem barely exists in ANZ right now.

Compliance is front of mind from conversation one. The Australian Privacy Act has been reformed. Automated decision-making transparency requirements come into effect in December 2026. Data sovereignty concerns are real and tightening. 72% of Australian companies now consider country of origin in AI vendor decisions, and over 80% see sovereign AI as a priority in strategic planning, according to Deloitte.

In regulated industries, the compliance question isn't a Phase 2 concern. It's a buying criteria.

A financial services client spent two months evaluating an AI platform before the data residency question came up and effectively restarted the procurement process. The US CLOUD Act, which allows American law enforcement to access data stored by US companies regardless of where it sits physically, creates real tension for Australian buyers with local data residency expectations. Any AI platform that doesn't have a compelling answer to this question will keep losing deals.

What's Actually Working

Despite the challenges, some companies are getting real results. The pattern isn't random.

They built a system, not a shortcut. Didn't just buy ChatGPT licenses and hope for the best. They identified one specific workflow, connected AI to their existing data and tools, and designed the human-AI handoff deliberately. The AI knew its role. The human knew their role. The process was documented.

One finance broker with 60 staff identified their proposal generation workflow as the bottleneck. They mapped the process, connected their CRM data, built a prompt framework, and cut average proposal time from two hours to fifteen minutes. 

They measured from day one. Specific metrics, set before the experiment started. Time saved. Output volume. Quality scores. Revenue impact. Two stats from the research sit side by side and tell the whole story: 82% of AI-using Australian businesses report a positive impact from AI, but 46% don't measure that impact at all, according to MYOB. And the Governance Institute of Australia found 93% of businesses lack effective ways to measure AI ROI in the first place.

You can't improve what you can't measure. The successful companies set their metrics upfront, killed experiments that didn't show results within 60 to 90 days, and doubled down on what worked rather than trying to do everything at once.

They invested in capability, not just access. The tools are table stakes. Everyone has access to the same tools. The gap is in how well teams can use them. The companies seeing the best results spent time on prompt engineering, workflow design, and building internal knowledge. They trained their people. This sounds obvious. It's widely ignored.

The implementation layer is where value is created or destroyed.

What This Means for AI Platforms Entering Australia

If you're building an AI platform's presence in Australia right now, what I'm seeing across the market has some clear implications.

You need a local partner ecosystem. Direct only won't scale in a market this fragmented, with buyers this pragmatic, and implementation capacity this constrained. Whether you're selling to a 50-person company or a 5,000-person enterprise, Aussie buyers need help going beyond the obvious use cases. They want local providers who understand their context, their compliance environment, and their operational reality. The AI platforms that invest in building this ecosystem early will win disproportionate share.

This is how HubSpot built its ANZ business. Not by selling directly to everyone, but by building a network of partners who could sell, implement, and support at scale. I watched that play out from the inside for 5.5 years. The partners who were invested in early became the market. 

Lead with local proof, not global vision. ANZ buyers don't care about your Fortune 500 US logos. One CBA case study is worth ten enterprise references from companies they've never heard of. Build local proof points as quickly as possible. POC programmes that let companies test with their own data will convert faster than any demo, regardless of company size.

Treat data sovereignty as a Day 1 concern. The Privacy Act reforms coming in December 2026 will make this more urgent, not less. Financial services and government buyers need local data assurance now. Any platform that pushes this to Phase 2 will keep losing deals to platforms that don't. The government angle matters too: Anthropic has already secured approval to supply non-corporate Australian federal government agencies. That's a signal about where the serious growth is, and it requires serious compliance investment.

Hire people who've built vendor GTM in ANZ before. You can't run this market from San Francisco. You can't run it from Singapore either. The market rewards local knowledge, local relationships, and local presence. The platforms that have figured this out have hired accordingly. Anthropic's international leadership has deep ANZ relevant pedigree: Chris Ciauri, their MD International, grew Salesforce EMEA from $200 million to over $3 billion in revenue. Paul Smith, their CCO, scaled ServiceNow's worldwide field operations through $12 billion in annual revenue. These are people who've executed exactly this playbook before, in markets with similar dynamics. That matters.

The knowledge of how to build an ANZ market isn't in your US team. It lives in the people who've already done it here.

Where This Is Heading

The next 12 months will determine which AI platforms establish a real presence in ANZ and which remain US-centric brands with a handful of Australian logos.

The market is ready. Adoption is real and accelerating. But it's happening on ANZ terms, not Silicon Valley terms. Pragmatic buyers who want proof. Lean teams who need implementation support. Compliance requirements that are tightening. A partner ecosystem that's thin but growing fast.

The platforms that understand this distinction will win. The ones that apply the US playbook and wonder why conversion is slower than expected will keep scratching their heads.

I'll be watching this closely and writing about what I see. If you're navigating AI adoption in your business, or building an AI platform's presence in this market, I'm always interested in the conversation.

References

• Anthropic Economic Index / CNBC, September 2025
• Deloitte, State of AI in the Enterprise 2026 
• Expert Market Research, Australia AI Market Report 2025
• Google / Ipsos, Australian AI Adoption Survey 2025
• Governance Institute of Australia, AI Governance Survey 2025
• IT Brief Australia, Anthropic Australia coverage, January 2026
• MYOB Mid-Market Survey, October 2025 
• MYOB SME Survey, November 2025 
• CSIRO, AI Ecosystem Report
• Bird & Bird, Australian AI Regulatory Tracker 2025-26
• IAPP, Australian Privacy Act Reform Analysis 2025
• CNBC, Paul Smith / Anthropic enterprise commentary, September 2025