AI not yet tested for consumer protection: ExpertBY MATTHEW WAI | WEDNESDAY, 18 NOV 2026 11:42AMArtificial intelligence (AI) has the potential to turbocharge Australia's advice sector, but an expert warns the innovation has yet to be tested at a consumer protection level. Speaking on a panel at the ASIC Annual Forum 2026, Super Consumers Australia chief executive Xavier O'Halloran, Wealthadvice managing director Marissa Broome, and Vanguard Australia managing director, Asia Pacific Daniel Shrimski discussed the important role AI is playing for both consumers and financial advisers. Despite some positive outcomes, including 24/7 access to advice and reduced cost to broaden advice access for those that never "darkened the door" of an adviser, O'Halloran remained sceptical of traditional AI models available to consumers. "I still think the focus and emphasis need to be on how to make sure it doesn't have those errors in it... I don't think anyone is properly engaged in thinking about that now," he stressed. "We're not testing the consumer protections around AI now in Australia, and I think we should be. We should be thinking about regulating the models to make sure they're delivering good outcomes to consumers, and all the ways that we can help feed better data to it." Highlighting the issue in AI, Broome also noted some of her clients had come into their meetings with AI generated advice, which were "horrendously" out of date and wrong. "When you hear errors like that, it can be really concerning. At the same time, we are a country not just of low financial literacy, but low financial capability," she said. "Not only do we not understand the jargon we talk about as a financial services ecosystem, but we can't make decisions around it based on that jargon. I'm really concerned that we're not doing enough in that space." Shrimski believes this trend signifies the importance of placing a human in the loop amid the rapid AI buildout. "I don't know exactly where we will get to, but I think human judgment is going to be critical when there's a coding error," he said. "Maybe we can't see it, but people using some basis of judgment and understanding what they're doing, and not just directly taking whatever AI is giving them and acting on it. It's easier said than done... But human involvement is going to be important." However, O'Halloran highlighted AI can provide positive outcomes. Being part of the RG 97 review crew, he discovered most information being drawn by an AI model is based heavily on product disclosure statements (PDS), which is helping consumers to understand massive sources of information in a simple way. "That's the kind of thing AI will feed on and be good at sending out to people. I think there's an opportunity there. We've got this regulatory architecture based around disclosure, which has never really worked because no one reads the stuff. But now we've got these AI agents and LLMs that can read and synthesise and making it simple for people." The Council of Australian Life Insurers (CALI) substantiates the comments, with findings from its Life Insurance Sentiment Tracker indicating the rate of AI users receiving financial advice had increased 80% to nearly one in three life insurance advice recipients. CALI chief executive Christine Cupitt said a convincing answer from AI could give families false confidence. The research also highlighted the continuing unmet demand for advice, with an increased use of digital tools and people using online product comparison websites rising to 25%, up from 19%. "An AI agent can give you an answer in seconds, the consequences of getting it wrong last for decades," Cupitt said. "Australians have too much on the line to leave to an untested and unregulated AI agent to answer for reliable financial advice." The research found more than half of working Australians (55%) say speaking with a real person is very important when receiving financial advice. Echoing Shrimski's statement, Cupitt highlighted "a real person matters when the questions are personal". "Technology may make information easier to find, but people need to know whether that information is accurate, or even relevant to their specific needs," she said. In response, CALI welcomed the government's August announcement that it would proceed with a New Class of Adviser for life insurers and APRA-regulated super funds, as Cupitt warned consumers shouldn't need to turn to AI because help was "out of reach". "The government has recognised Australians needed more ways to access advice and Australia's advice system must keep pace with how people are seeking help," she said. "Australians should be able to ask their life insurer a simple question and get a simple answer. They shouldn't have to turn to AI agents because reliable help is out of reach." Related News |
Editor's Choice
ASIC reveals 2027 enforcement priorities
WAM's halved FY27 dividend at risk as investment losses mount
Praemium confirms C-suite appointments
ART pushes for expansion of intrafund advice scope
Products
Expert Feed
Featured Profile

Fiona Macgregor
TAL LIMITED






