A/NZ tech leaders warn AI governance lags adoption

Technology leaders in Australia and New Zealand are using this year's AI Appreciation Day to warn that governance and clarity of purpose are lagging behind rapid adoption. Executives from F5, ManageEngine and Constant Contact say many organisations are moving faster on deployment than on discipline.

The comments underline how AI use in the region has shifted from experimentation to execution over the past year. Large enterprises and small businesses are now running production systems, even as concerns over risk, data foundations and customer trust intensify.

Jason Baden, Regional Vice-President A/NZ at F5, said many boards now treat visible progress on AI as a priority metric. That, he said, has fuelled a rush into pilots and proofs of concept without enough upfront thinking about the problems they are meant to solve.

"2025's AI Appreciation Day saw businesses debating how and when to experiment with AI. This year, the conversation looks very different. In many cases, the tools are deployed, the pilots are done, and the real test has begun. After a year spent speaking with executives, technologists, and customers across A/NZ, the enthusiasm hasn't surprised me. What has surprised me is how rarely anyone pauses to ask the most basic question before committing budget and resources. Not 'can we do this?' but 'should we do this, for this problem, right now?' That discipline - knowing what you're solving before reaching for the tool - is rarer than it sounds, and far more valuable. I've watched organisations spin up proof-of-concept projects, hit a wall, and struggle to understand why. The answer is almost never the technology. It's that nobody defined the problem clearly enough before the project started. That's exactly what's happening in boardrooms and strategy offsites across the region, where the pressure to demonstrate progress has become its own kind of risk. The business wants to move quickly, and use cases are appearing everywhere. I understand that pressure and I feel it myself. But speed without clarity is how organisations end up with expensive infrastructure, unsettled people, and governance functions scrambling to catch up with decisions already made. If there's one thing worth appreciating on 2026's AI Appreciation Day, it's not the models but the foundation underneath them. Most AI initiatives don't stumble because of the technology. They stumble because the data feeding it isn't governed or even understood properly. That's the harder work nobody wants to celebrate, but it's the work that will actually determine whether AI delivers," Baden said.

Baden's comments reflect a broader concern among technology leaders about data readiness. Many see gaps in data quality and ownership as a bigger obstacle than algorithm performance.

Risk and security leaders are also focusing on the emergence of so-called agentic AI systems, which can take actions, make decisions and adapt workflows without step-by-step human instruction.

Vinayak Sreedhar, Country Head A/NZ at ManageEngine, said the rise of these autonomous agents is changing the security conversation inside IT teams. Organisations, he said, must treat each autonomous system as a distinct identity that can move through networks and trigger changes.

"AI Appreciation Day is here again and this year, it's worth remembering AI isn't just generating content or analysing data anymore - it's beginning to make decisions, trigger actions, adapt in real time, and automate complex processes. The move toward agentic AI is changing how organisations manage technology, risk, and trust. But it's also exposing a gap in how many approach security. Traditional automation was predictable. Agentic systems analyse multiple variables, weigh trade-offs, adjust course, and decide what action to take. That's worth celebrating, absolutely. But autonomy brings a completely new category of risk. When machines act on an organisation's behalf, identity and access governance becomes mission critical. Every autonomous system is effectively a new digital identity inside the organisation. One capable of accessing systems and executing workflows at that. Over-permissioned agents don't just speed deployment, they dramatically increase the attack surface. The answer is 'safe autonomy' where agents operate within human oversight, tightly defined boundaries, clear permissions, and constant monitoring. So today isn't just about appreciating what AI can do. It's about appreciating the discipline required to deploy it responsibly. The question isn't whether to deploy AI. It's whether it can be deployed safely. As machines take on more decisions, governance over identity and access will be what protects trust," Sreedhar said.

While large enterprises wrestle with governance and security, smaller firms in A/NZ are ramping up day-to-day AI use in marketing and customer engagement. Adoption is now widespread among small and mid-sized businesses.

Renée Chaplin, VP - Asia Pacific at Constant Contact, said AI has quickly become embedded in routine marketing tasks for smaller firms in the region. She pointed to new research from the company showing high uptake and largely pragmatic motivations.

"For many small businesses, AI has become the extra set of hands they never had the budget to hire, and it's exactly why adoption has taken off so quickly across Australia and New Zealand. Constant Contact's latest Small Business Now report (https://www.constantcontact.com/blog/small-business-marketing-statistics/) revealed that across the region, AI adoption in SMB marketing has reached 88.7 per cent, the highest rate of any surveyed region. It's clear AI use has moved well beyond early adopters into everyday practice. What's driving it is completely practical. More than half of ANZ SMBs using AI point to saving time as the biggest win, whether it's drafting emails and subject lines, analysing data, or creating visual content. For a business owner wearing every hat at once, those reclaimed hours are incredibly valuable and can go straight back into customers and growth to drive the business forward. That said, the shift comes with some real concerns. Many SMB owners fear AI use could lead their customers to feel a loss of personal touch, or raise privacy, authenticity and trust issues. Addressing these concerns boils down to how tools are used and for what purpose. Used thoughtfully, AI should deepen the human connection, not dilute it," Chaplin said.