[GRTV] FMA puts AI advice and banks’ return to wealth management under spotlight
Watch FMA Director of Deposit Taking, Insurance, and Advice Michael Hewes regulating AI-generated financial advice and why banks should consider getting back into the wealth market.
Monday, July 27th 2026, 8:37AM 
2 Comments
Artificial intelligence and the prospect of banks moving back into financial advice are emerging issues for the Financial Markets Authority following its landmark Access to Advice research.
In the second part of a GRTV interview with FMA director of deposit taking, insurance and advice Michael Hewes, the regulator discusses the growing use of AI by consumers and advisers, vertical integration and what happens next following the report.Hewes says AI will be a particular focus for the FMA over the coming year as it builds its own capability and works with industry organisations.
The regulator sees AI as a potentially valuable tool for advisers, but Hewes warns about the technology’s tendency to produce incorrect information, or “hallucinations”.For consumers using AI directly for financial guidance, the message is to test its recommendations and, where appropriate, put a “human set of eyes” over the results.
Hewes says regulators internationally are grappling with the same challenge. The FMA is looking at work being done by overseas counterparts, including Australian and UK regulators, as it considers what guardrails may be needed.The interview also explores whether banks and other vertically integrated financial institutions could play a bigger role in advice.
FMA research found bricks-and-mortar institutions remain important to consumers, while Hewes says it can sometimes be easier for institutions not to provide advice. That creates an opportunity for banks and other providers to reconsider their advice models.However, any return to advice would take place in a different regulatory environment from that which existed before Australia’s Royal Commission.
Hewes points to the Conduct of Financial Institutions regime, fair conduct programmes and stronger expectations around consumer outcomes as safeguards against some of the problems seen previously.He also confirms the FMA expects vertically integrated providers to refer customers to external advisers where appropriate and says some banks already have arrangements with specialist insurance advisers.
Hewes says the Access to Advice report has been positively received by the advice sector and suggests the FMA could revisit the research in about 24 months to measure progress.Watch the full GRTV interview with Michael Hewes for his views on AI advice, banks returning to wealth management and what the FMA plans to do next.
| « [GRTV] How to get better access to advice |
Special Offers
Comments from our readers
I think we need a thematic on thematics. I think most thematics are academic and a waste of money.
The advice paper referred to was not revolutionary and in summary indicated adviser/FAP terror on giving or creating advice. This led to overcooking advice making it expensive to provide which led to (in my mind) some segments not getting service or less people getting service as the expensive baking takes too long.
Some segments are unfortunately not reliable to service and have risk of default to an adviser. AI can help in time with those perhaps.
We can, "unpack" eco-systems and take snapshots whilst considering "guardrails". We could consider VIO's and create "benchmarks" and hope for better consumer outcomes or... we could just get on with it.
I dont imagine trips to Aus or the UK to consider how their regulators are pondering AI are that useful. Good junkets and count me in but in reality, this is not complicated. Data security and advice accuracy are the key factors in an AI model for advice.
Sign In to add your comment
|
|
Printable version |
|
Email to a friend |
Personally, I think any financial institution or adviser should be free to use whatever tool available, including AI, to assist them. However, they will have to be ultimately responsible for their recommendation, ie. cannot blame the tool or AI if something goes wrong.
However, if an investor / a consumer chose to rely on AI, that's his prerogative. Shouldn't blame anyone if something goes wrong.