Katherine Wilson, CEO, of Insurance Brokers Association of New Zealand (IBANZ), explores the implications of using AI for insurance advice.

It’s natural for investors to seek an edge.

They read prospectuses, analyse fund disclosures, compare fee structures and interrogate performance data. Increasingly, they are also turning to generative AI tools to synthesise information, test assumptions and explore investment strategies.

It’s a very rational and reasonable approach. AI tools can summarise complex topics in seconds and effortlessly compare features across multiple financial products. Used carefully, they can function as highly efficient research assistants.

But there is a line that’s becoming increasingly blurred.

In a recent address to more than 450 IBANZ members, Chapman Tripp partner Tim Williams raised a concern that deserves wider consideration. Under New Zealand’s Financial Markets Conduct Act (FMCA), offering financial advice is a regulated activity.

Providing factual information is ok, but recommending a specific product or constructing a personalised investment strategy crosses the line into providing financial advice.

No AI chatbot is licensed as a financial advice provider. So, when an AI tool recommends a specific insurance policy or identifies a particular exchange-traded fund as the “best” option, it is no longer simply providing information – it is straying into the territory of regulated financial advice and that is in breach of New Zealand law as it currently stands.

Investor protection

This is not about semantics. It goes to the heart of how investor protection operates in New Zealand. Giving financial advice is a regulated activity for good reason.

Qualified financial advisers must meet minimum competence standards, including completing the New Zealand Certificate in Financial Services (Level 5). That qualification ensures advisers understand regulatory obligations, ethical duties, client suitability requirements, product structures and the real-world consequences of advice.

Beyond qualification, advisers must operate under a licensed financial advice provider, comply with the Code of Professional Conduct, undertake continuing professional development, manage conflicts of interest, and remain accountable to the Financial Markets Authority (FMA).

This framework exists because financial decisions carry risk. It reflects a public policy position that influencing another person’s financial decisions in a personalised way should sit within a system of training, standards and accountability.

When clients receive advice from a licensed financial adviser such as an insurance broker they are engaging with someone who is trained, supervised and legally responsible for their recommendations.

When a recommendation comes from a generative AI tool, none of those safeguards apply. There’s no qualification, no code of conduct, no professional supervision, no regulatory accountability and no formal avenue for redress.

Significant concern

A recent IBANZ member survey nominated the risk of unregulated AI advice as the single highest advocacy concern for members. It ranked above legislative reform, levy concerns, affordability issues and other significant regulatory matters.

Insurance brokers operate at the front line of financial risk management. They see the consequences of poorly structured cover and understand how policy wording differences can materially affect claims outcomes. Concern about AI-generated advice reflects what they are seeing in practice: consumers increasingly relying on AI tools to make significant financial decisions.

For example, using AI to calculate insurance risk may seem straightforward, but the real complexity lies in the policy wording – and those same principles don’t necessarily translate across regions, varying demographics, or when assessing broader business or commercial property risks. Put simply, AI doesn’t give personalised advice, ask the right follow-up questions, or make sure your cover truly matches your needs.

AI outputs often feel authoritative. They are structured, confident, they cite data and appear logical. For active investors accustomed to disciplined analysis, this sort of presentation can carry weight.

But authority is not accountability

AI systems do not have a statutory duty to act in a client’s best interests. They are not required to conduct suitability assessments in accordance with FMCA requirements. They don’t document advice, manage conflicts, or have an obligation to meet professional standards.

Insurance is particularly vulnerable to oversimplification. Policy definitions, exclusions, underwriting criteria and the interaction between personal and business covers are rarely straightforward. Seemingly insignificant variations in wording can result in underinsurance at claims time.

Insurance brokers must hold the Level 5 qualification and meet ongoing competence standards precisely because these decisions require context, judgement and accountability. They are trained to assess client needs holistically, explain trade-offs, and document recommendations in line with regulatory requirements.

Place for AI

Of course AI has a place. But it is not regulated advice and the two should not be confused.

This issue it not confined to New Zealand, it’s already being noted as a problem in jurisdictions such at the United Kingdom and Australia where consumer groups and advisers are highlighting how free AI tools are making share-trading recommendations and providing personalised financial-planning advice.

When financial futures are at stake, advice should come from trained professionals operating within a defined regulatory framework. Technology will no doubt continue to evolve; New Zealand regulators will have to join their international colleagues in considering how to respond in Kiwis’ best interests.