When Amazon Web Services, more commonly known as AWS, planted their flag on Australian soil 13 years ago, cloud computing was still emerging. Today, it powers almost everything we do day to day; consumer banking, enterprise platforms as well as systems tied to national security. AI adoption accelerates across Australia and New Zealand, the question is no longer whether the cloud can scale fast enough, but whether trust, resilience, and governance are keeping pace.
Recently, I attended a closed door ANZ press briefing with AWS Australia and New Zealand Vice President and Managing Director, Rianne Van Veldhuizen, and Director of Solutions Architecture APJ, Nam Je Cho at their annual re:Invent conference in Las Vegas.
“[The] acceleration of cloud adoption across customers and also the acceleration of consumption with AI is something that we’ve definitely seen in Australia and I’m personally very proud of because … it’s moving a lot quicker in Australia than I see in some of the countries where I’m originally from,” Van Veldhuizen shared.
Van Veldhuizen framed AWS’s local journey as a shift from startup enablement to national infrastructure. Early partnerships with companies like Atlassian and Canva have given a leg up to deep public sector engagement and a reported $20AU billion investment in Australian infrastructure. Australia, she explained is adopting cloud and AI faster than many comparable markets, even though Australia is perceived as a reserved market.
AWS’s workforce and training investments are part of that strategy. Since 2017, the company has trained more than 400,000 people locally and plans to expand AI education initiatives, including programs aimed at K–12 students. The goal is to make AI adoption mainstream, accessible and economically meaningful.
“We’ve [AWS] trained over 400,000 people on skills since 2017. And that’s not going to stop because you can imagine that with AI, if we’re talking about everybody thriving with AI, then there is an opportunity for all of us to level up. That includes everybody at AWS as well, but also everybody at our customers,” Van Veldhuizen went on say.
Alongside this push came a slate of AI announcements which included sovereign AI infrastructure, new multimodal models and autonomous agents. Together, they reflect customer demand for flexibility across regulation, performance and cost. Above all security is still takes the cake for the hyperscaler.
“Security is and will always continue to be our number one priority and number one job,” Cho added.
Customers nowadays need and want choice, not a single dominant locked in model. This way of thinking is where Bedrock comes in – this is the breeding ground for where tens of thousands of organisations are experimenting with multiple foundation models.
But Cho was also up front about failure and that no technology company out there has the silver bullet. Cloud systems are physical systems, subject to outages, dependencies and human error. Mr Cho quotes AWS CTO Werner Vogels “everything fails all the time” he reinforced that resilience is not about avoiding failure, but designing for it.
“I’ll go back a few years when Werner’s said some wise words, where everything fails all the time. And one of the things that I think we focus on massively is around our operational excellence,” said Cho.
Recent outages reiterated that point. While most Australian customers were unaffected, some US hosted workloads experienced disruption. AWS’s view is that hyperscale platforms still offer resilience beyond what most organisations can achieve on-premise, but only if customers actively architect for failure.
“You have to assume things will go wrong,” both executives acknowledged.
AWS cites strong indicators of AI momentum and the majority of Australian startups now begin with AI front of mind. Heavily regulated sectors like financial services are moving faster than expected given the nature of their complexity and red tape. Van Veldhuizen noted that banks, once cautious were able to accelerate adoption because foundational controls were already in place.
“[Fifty] percent of the Australian businesses have adopted AI and are either in experimentation or are driving that in production. And we see that growth accelerating where maybe two years ago it was mostly experimentation and we’re now beyond that point. Customers want to see a clear ROI,” Van Veldhuizen said.
AI systems that operate at scale, particularly autonomous or agent based systems can amplify the challenge of potential oversight. Even with guardrails such as memory, policy enforcement and continuous monitoring, complexity grows faster than auditability.
Cho outlined AWS’s efforts to address this gap by embedding context, enforcing consistency and monitoring agent behaviour in real time. These are necessary steps. Whether they are sufficient as AI systems become more independent remains an open question many large technology vendors have.
AWS leaders were forthright in the sense that technology alone is not the solution, but is part of it. Resilience depends on shared responsibility which encompass customers, partners and platform providers working continuously to test assumptions and review architecture. Training, briefings and partner engagement are central to AWS’s approach.
