Consumers Face Risks With AI Money Advice as FMA Surveys Industry Adoption

New Zealand consumers using generative AI tools for budgeting, investment picks and financial queries risk generic or inaccurate recommendations not tailored to local conditions.

A podcast test showed ChatGPT recommending Rakon shares to a New Zealand investor shortly before the company delisted from the NZX. Kernel founder Rupert Carlyon described it as one of the worst NZX options at the time.

University of Auckland Professor Alexandra Andhov said AI outputs remain generic and potentially biased toward advertised products. They lack personalisation for New Zealand-specific circumstances such as tax rules or local market conditions.

FMA Survey: Industry Moving Fast

The Financial Markets Authority surveyed 30 regulated entities in May and June 2024. Thirteen responded, a 40 percent rate across banking, insurance, advice and asset management sectors.

Nine of the 13 respondents already used AI technologies. All planned further integration of generative AI or machine learning.

FMA spokesperson Romil Ghelani said digitisation and AI have made financial information more accessible. New Zealand's regulatory framework is technology-neutral, meaning the same legal obligations apply regardless of whether advice is delivered digitally or in person.

Ghelani added that AI tools and global platforms generally provide financial information rather than regulated advice. Financial information includes general data, statistics or insights without recommendations on specific products.

"The FMA is technology-neutral and pro-innovation. We believe that New Zealanders should have access to the same technological advancements as those in other countries." — Stuart Johnson, FMA Chief Economist

Firms in the survey reported realising efficiency benefits already or within 12 months. Most maintain human oversight, staff training and risk frameworks to address hallucinations and bias.

RBNZ Flags Stability Risks

The Reserve Bank of New Zealand noted in its May 2025 Financial Stability Report that banks adopt AI faster than insurers due to competitive pressures. Reluctance persists for customer-facing generative AI because of risks and costs.

RBNZ highlighted benefits including improved productivity and risk assessment accuracy. Risks include model bias, data privacy breaches and market concentration on third-party providers such as NVIDIA.

Regulatory Framework and Consumer Recourse

FMA Chief Economist Stuart Johnson said the regulator is technology-neutral and pro-innovation. New Zealanders should have access to the same technological advancements as those in other countries.

An April 2026 FMA and University of Auckland ALTeR Centre workshop drew over 40 stakeholders from banking, insurance, fintech and law. Discussions covered implications for consumer protection, transparency and accountability.

Licensed Financial Advice Providers remain responsible for AI-assisted outputs under the Financial Markets Conduct Act. Suitability requirements, best interests obligations and disclosure rules apply equally to digital and human delivery.

Unlicensed global platforms fall outside these protections. Consumers retain direct recourse through the Banking Ombudsman Scheme, which has issued guidance on verifying AI-generated complaint content.

Regulators continue monitoring through domestic and Trans-Tasman engagement. Ongoing roundtables will shape future guidance on governance and explainability.