New Zealand’s financial advice sector is operating in a regulatory gap that insurance brokers cannot afford to ignore. AI tools used to support client advice sit increasingly outside the licensed framework, professional indemnity (PI) policies are beginning to exclude AI-related errors, and the regulator is seeking to build its first comprehensive picture of what is happening on the ground. That is the backdrop to the Financial Markets Authority’s (FMA) launch of a thematic review into artificial intelligence in financial advice, announced August 6, 2026. Four targeted surveys are open to industry participants until September 4, 2026.
The risk is not a regulatory abstraction. A recent Insurance Brokers Association of New Zealand (IBANZ) member survey nominated the risk of unregulated AI advice as the single highest advocacy concern for members – ranking above legislative reform, levy concerns, affordability issues, and other significant regulatory matters. IBANZ CEO Katherine Wilson, writing in Informed Investor in May 2026, explained what is driving that concern: insurance brokers operate at the front line of financial risk management and see the consequences of poorly structured cover. AI outputs often feel authoritative – structured, confident, citing data – but AI systems do not have a statutory duty to act in a client’s best interests, are not required to conduct suitability assessments, and have no obligation to meet professional standards.
Chapman Tripp partner Tim Williams demonstrated the problem live to over 450 IBANZ members earlier this year, showing how popular AI tools recommended specific life insurance policies and named investment funds. His conclusion, as reported by b2bnews.co.nz, was direct: when an AI tool moves into recommending a specific product, “it is clearly giving investment advice, and that is in breach of New Zealand law as it is currently written.”
The regulatory framework under the Financial Markets Conduct Act 2013 is technology-neutral, but the FMA’s own engagement overview concedes its practical application “may become more complex where AI supports, shapes or generates part of the advice process.” The Code of Professional Conduct for Financial Advice Services Version 2, which came into force November 1, 2025, establishes nine standards covering suitability, client understanding, and information protection – with no AI carve-out. An earlier regulatory accommodation, the Financial Advisers (Personalised Digital Advice) Exemption Notice 2018, expired on May 31, 2023, leaving a vacuum precisely as generative AI proliferated.
Chapman Tripp has noted that many AI tools are operated by entities that are not licensed and regulated by the FMA, yet often still provide financial advice on prompting from a user, effectively denying clients the full protections of the law. Buddle Findlay’s January 2026 analysis found that 82% to 87% of Kiwi organisations use AI but only 34% of New Zealanders trust it – warning that a high-profile AI advice failure reaching the courts could cost significantly more in lost business and reputational damage than the lawsuit itself.
The FMA’s enforcement record shows it acts on conduct failures. Westpac was ordered to pay $3.25 million and AA Insurance $6.175 million in penalties for fair dealing breaches in the 2024/25 year. FMA chief executive Samantha Barrass, speaking at the Financial Advice New Zealand (FANZ) Conference in March 2026, acknowledged technology as central to closing the advice access gap – noting that it “can automate parts of the process that currently take advisers a lot of time, freeing them up to focus more on the human conversations that consumers value” – but the regulatory framework has not yet caught up with what is already in the market.
For brokers, the most immediate commercial implication of the FMA’s review may be what is already changing in the PI market. Under the Financial Markets Conduct Act 2013 and the Code of Professional Conduct for Financial Advice Services, financial advice providers (FAPs) must hold PI insurance or demonstrate equivalent arrangements. MinterEllisonRuddWatts partners Andrew Horne and Tom Maasland, writing in February 2026, noted that NZ insurers are increasingly concerned about professionals placing reliance on AI-generated work without adequate human oversight – and predicted that insurers are likely to follow the cyber insurance model, asking increasingly detailed questions about AI governance, security, provenance, and regulatory awareness, “with potential impacts on premiums, exclusions, and coverage availability.”
That trajectory is confirmed globally. Gallagher Re’s report Smart Systems, Blind Spots: Rethinking Insurance for the AI Era, developed with MIT, found that the pace of AI adoption has outstripped the insurance industry’s capacity to develop responsive products, leaving organisations exposed to a growing class of uninsured liabilities. One in five insurance professionals reported a client had experienced an AI-related loss during the previous year, with just over half of those losses fully covered by insurance.
A May 2026 analysis by US law firm Honigman, referencing the Gallagher Re/MIT study, highlighted growing uncertainty around how existing insurance policies respond to AI-related losses. The firm warned that some insurers may argue certain AI risks were not contemplated when policies were written, potentially leading to disputes over coverage, exclusions, and policy interpretation. For brokers, the issue reinforces the importance of discussing clients’ AI use and reviewing whether existing coverage adequately addresses emerging exposures.
The FMA has said that depending on survey findings, it may publish key themes, engage further with industry on good practices, or use the results to inform regulatory guidance or future policy work. MinterEllisonRuddWatts noted in May 2026 that the FMA's regulatory sandbox has already delivered first-of-their-kind exemptions, signalling genuine intent to enable compliant innovation alongside oversight.
For brokers, the commercial stakes of each outcome differ. Formal regulatory guidance would give FAP clients a clearer compliance framework, reducing the ambiguity currently deterring AI adoption. No guidance leaves liability unresolved and PI coverage questions unanswered at renewal. Enforcement action against an unlicensed AI advice tool would, as Buddle Findlay warned, likely cost any connected business far more than legal fees alone. Wilson put the broker’s role plainly: “Strong brokers are the ones that are future-ready. They’re not just focused on today’s challenges; they’re already anticipating what's coming next and planning how to respond.” The four surveys – directed at FAPs and advisers, technology service providers, legal advisers, and industry associations – close September 4, 2026.
