For funds, the findings are clear: AI’s greatest near-term value isn’t in replacing the adviser. It’s in supporting them, drastically scaling their impact to reach the people who need it most.
The Australians who most need retirement advice are the least likely to accept it from AI. PwC research finds that 68% of respondents aged 61-79 said they would not use an AI-powered tool for financial advice. Among 18–28-year-olds, that figure was 19%.
AI still has a role to play in serving older members. However, AI investment decisions need to be made segment-by-segment, and must be grounded in behavioural evidence, to genuinely reduce the advice gap.
What does this mean for the AI strategy decisions your fund is about to make?
What the upcoming advice reforms do and don't solve
Australia’s superannuation system is mid-shift, from accumulation engine to retirement income platform, and advice is the new competitive frontier.
The regulatory environment is responding. The Delivering Better Financial Outcomes (DBFO) reforms currently before Parliament will give trustees considerably more scope to provide personal advice on retirement planning, including the ability to consider household circumstances, debts, and Age Pension eligibility. Tranche 2 of these reforms proposes allowing personalised nudges at key life stages without triggering full personal advice compliance. A new ‘qualified adviser’ category, combined with a shorter advice record (replacing the Statement of Advice), would further reduce compliance cost per interaction.
Industry bodies have welcomed the regulatory clarity; however, the legislation is moving slowly, and retirees will not wait. Funds that fail to address the need for financial advice will see continued outflows to retail platforms, where advice is better embedded with the product.
Together with advancing AI, these regulatory changes make it economically possible to have a higher-volume, lower-cost human advice model. Advisers can handle a greater volume of better-prepared conversations supported by AI. Members still get a human adviser, but AI is bringing the cost per interaction down.
The case for human-led, AI-assisted advice
The question is not whether to invest in AI, rather it is where and how. Funds must make well-informed decisions about where AI faces the member and where it works behind the scenes to make human advice reach further and cost less.
AI's greatest contribution to the advice gap, at least in the short term, is in making human-led advice fast enough, relevant enough, and affordable enough to reach the people who need it most.
Keep the human adviser. Use AI to scale their impact.
In fact, our AI Performance Study reinforces this, highlighting that 20% of organisations capture 74% of AI-driven value, and achieve gains 7.2 times higher than peers, driven by a more deliberate focus on where AI is deployed.
Where does AI fit for Australian superannuation funds?
What fund members told us
We surveyed more than 3,100 Australians about their likelihood of using AI for financial advice, whether they had used public AI tools, and whether they would refuse AI for financial advice under certain circumstances. Responses were cross-referenced against age, superannuation balance, retirement planning status, and financial confidence.
The results do not describe a single adoption curve.
Perhaps unsurprisingly, AI receptivity declines sharply with age, but age is not the only variable. Respondents with a retirement plan, even a general one, were substantially more open to AI-assisted advice than those who had not thought about retirement at all. People’s super balance also showed an impact. Members with small-to-medium balances ($25,000–$199,999) showed the strongest openness, while those at the extremes, particularly those who could not or would not state their balance, were most resistant.
Gender made almost no difference. Given how much of the financial services industry's segmentation defaults to gender as a primary variable, that finding is worth noting. On AI receptivity, we found that age, financial confidence, and retirement planning status all matter more.
Some portion of the resistance captured in this survey will soften as AI becomes embedded in more everyday consumer experiences. But normalisation takes time, the US experience with robo-advisers, where early scepticism gave way to mainstream adoption, is instructive.
Four segments, four different responses
Digital-first adopters
Who? Young, already use AI tools, report high financial confidence and some retirement planning.
What? Trust in AI for this cohort is already established, making the operational task more straightforward. Advice needs will grow more complex over time.
Where AI fits? Deploy member-facing AI advice tools and ensure content keeps pace with needs. The risk this segment poses is not execution but misallocated prioritisation. Their enthusiasm makes AI investment cases look compelling in board papers, but they are the cohort with the longest time horizon, and the least complex advice needs right now. Funds that prioritise this segment first will build capability that serves the members who need it least. This capability will quickly become outdated as the technology and regulatory environment evolves.
Your next move? Member-facing AI, constantly updated.
The persuadable middle
Who? The largest segment at approximately 38% of surveyed sample, and the one most open to influence, typically aged 29–44. Willing but untested, they score between ‘likely’ and ‘maybe’ on willingness to use an AI advice tool, but almost none have actually used one. This segment is the highest-priority design challenge in AI advice.
What? Trust in AI is yet to be developed, making the first few encounters crucial. For retirement and financial services, well-designed initial interactions, personalised, simplified, with a clear next step, generate materially higher sustained engagement than those built around information volume or functionality alone.1
Where AI fits? Member-facing AI is right for this segment, but it must be thoughtfully designed. An AI advice tool that opens with a personalised, concrete picture of a member's retirement position, and signposts the next step clearly, will outperform one that leads with complex features or generic information. The advice regulatory landscape has historically hampered funds from leaning into proactive member engagement such as nudges, with many choosing to wait for regulatory clarity.
Your next move? Invest now in user testing, behavioural research, and iterative prototyping for this segment's first AI touchpoint.
The disengaged
Who? Characterised less by hostility to AI and more by a lack of financial engagement altogether. Low planning status, inability to state their superannuation balance, and high AI refusal rates are more likely to be a default position rather than a considered judgement. This is not a hypothetical population. When APRA conducted its first MySuper performance test in 2021, 13 products failed, collectively holding approximately 1.1 million member accounts.2 Those members received direct warning letters. Only a minority moved. Most stayed where they were, despite a direct prompt.
What? Telling disengaged members their outcomes are poor does not, on its own, change behaviour. What works follows a clear hierarchy:
- Structural interventions, changes to defaults and commitment mechanisms produce the largest effects. The UK's introduction of auto-enrolment in 2012, underpinned by the National Employment Savings Trust (NEST) as the default scheme, lifted workplace pension participation from 47% of UK employees in 2012 to 76% in 2018,a 29% increase achieved primarily by changing the default rather than the disclosure.3
- Personalised guidance that addresses individual circumstances moves behaviour measurably. When participants in a fund choice experiment were given access to a smart calculator that computed fees based on their own balance, understanding improved significantly, whereas a basic calculator without that personalisation had no effect for most users.4
- Generic disclosure consistently fails. Long, complex communications that restate regulatory information without simplifying the decision do not shift outcomes.
Most structural levers sit with regulators. What funds can control is the design of their communications. A generic retirement projection, sent to a member who has never opened a statement, isn’t likely to have an impact.
Where AI fits? Engaging this segment requires personalisation, simplification and putting a specific, low-friction next step in front of members. For example, a pre-booked guidance call or a single-click action. Without significant design effort upfront, the nudge provisions in tranche 2 of DBFO will just produce more communications that the disengaged will continue to ignore. AI has the potential to accelerate the design, implementation, monitoring and evolution of nudges, but care is required to ensure these nudges deliver the right outcomes for members.
Your next move? Use AI to personalise communication, starting with one specific, low-friction step.
Advice-open, AI-resistant
Who? The most strategically significant segment for the near term. Older members, often with moderate to strong retirement planning engagement, who want advice but have actively decided against receiving it through AI. They do not trust AI to handle something as consequential as superannuation, and that judgement is unlikely to shift through better marketing or interface design.
What? This segment needs human-led advice at the life stage where advice is most complex and most expensive to deliver. The paradox is urgent because it is temporary. Whether the resistance in this oldest cohort is purely generational is worth asking. Today's 45–60-year-olds are substantially more digitally fluent and will likely bring different attitudes when they reach retirement. The long-term trajectory favours broader acceptance but leaves the question – what do funds do for people retiring today?
Where AI fits? AI's role here is to restructure the cost base behind human interactions. Think: preparation tools that assemble a member's financial picture before the conversation; triage systems that identify members approaching decision points and route them to advisers at the right moment; scenario modelling that gives an adviser pre-built retirement income projections to walk through rather than building each from scratch.
Your next move? Meet members where they are. Design the advice conversation so AI does the preparation, and the adviser does the talking.
The investment decision
Delivering across four different engagement models will require partnerships with: technology providers building adaptive interfaces; financial advisers delivering human capability at volume; behavioural specialists designing evidence-based interventions for the hardest-to-reach; and, over time, government agencies whose data and entitlements are central to retirement decisions.
The members who most need help are the ones who will only act if they trust the advice experience.
How funds respond will shape whether AI-enabled advice reaches those who currently go without or simply shortens the wait for those already in line.
This is the second in a series of articles examining how superannuation funds can turn scale into value for members. This follows our first article, ‘Value in Motion: $4.5 trillion in super savings. Now what?,’ which explored the industry's transition from accumulation to decumulation and the growing importance of advice and member engagement.
Value in Motion: $4.5 trillion in super savings. Now what?
Australia has built a world-class system for growing retirement savings. But is it ready for what comes next?
Value in Motion: Financial Services
Within a decade, the Financial Services industry will be redefined by new needs – and new players. With trillions in value on the table, Australia’s capital stewards will need to evolve to lead what comes next.
Contact us
