AI Pulse · ANZ AI Watch

Australian and New Zealand organisations are moving AI agents out of pilots and into everyday work quickly. Salesforce’s 2026 Connectivity Benchmark reports that 89% now deploy them across most or all teams. The gap is that only 54% have a formal governance framework in place.

That matters because an AI agent is not just another software licence. It can access customer records, draft communications, trigger workflows, and influence decisions at scale. If nobody has set clear rules for access, approval, monitoring, data handling, and accountability, the business is taking on risk without a clear owner for it. The same research shows 83% of Australian service professionals say regulation is actively constraining deployment, compared with 68% globally. That suggests many teams are already feeling the friction of trying to put controls in place after the tools have arrived.

New Zealand has its own split. It trails on governance, but leads on measurement, with 84% tracking AI adoption rates, 20 points above the global average. Tracking usage is useful, but it is not the same as knowing whether agents are acting safely, producing reliable outputs, or accessing the right information.

For owners, the practical question is simple. Can you name the person accountable for each AI use case, the data it can access, and the checks that happen when it gets something wrong? If not, adoption may be ahead of management. This is the kind of thing we build into an AI command centre, bringing usage, ownership, controls, and business impact into one operating view.