_svg.webp){kind=logo}
AI, Human Support and the Future of Financial Planning
Vanguard’s The AI advice frontier: Use, trust, and the human edge explores how 6,686 investors use AI and what they still want from human financial professionals. Most use AI to learn, explore ideas and prepare for decisions, while remaining cautious about accuracy and reluctant to delegate control. The findings suggest that human support retains considerable value—but leave open how that value will be delivered and paid for as clients become more capable. For financial planners, the question is whether AI strengthens the existing advice model or creates demand for a different relationship: clients owning their plans, with human expertise available when needed.
This supports the Academy’s model more strongly than its headline suggests—but it does not prove that the existing advice business model is protected.
Vanguard’s The AI advice frontier surveys 6,686 Vanguard investors about AI use, trust and preferences. Its central finding is that people want AI to help them understand and explore, while retaining human support for consequential decisions. the_ai_advice_frontier.pdf
1. The strongest finding is about retaining control.
Among AI users, 54–61% use it as a starting point for learning, and 48–61% as a sounding board. Only 1–4% use it to make decisions on their behalf. Across generations, 80–85% are uncomfortable with AI acting independently.
That is remarkably close to your proposition: help people understand, choose and act themselves, with expertise available when needed.
People may want more intelligence without wanting to surrender more authority.
2. “More valued” does not mean “the same business model survives”.
The report finds that 32% value human advice more because of AI, against 2.9% who value it less. That is useful evidence of perceived value. But the headline—“AI is increasing demand for human advice”—goes further than the survey establishes.
It does not demonstrate increased purchases, willingness to pay, adviser revenue or acceptance of ongoing asset-based charges.
A person could value a skilled human more at a difficult decision point while needing fewer routine services overall. Higher value per conversation can coexist with fewer paid conversations. That is entirely compatible with continuous capability and episodic expertise.
3. The human advantage needs defining more carefully.
The report stresses connection, empathy and understanding someone’s circumstances.
Those are valuable, but they do not establish that an adviser must continuously manage the relationship, plan or assets.
For the Academy, the useful question is:
Does human support leave the person better able to navigate the next decision?
Reassurance can build capability—or sustain dependency. The survey does not distinguish between those outcomes.
4. The trust gap gives the Academy a concrete job.
Among AI users, 37% report receiving incorrect or misleading information, while 44% cannot tell whether they have. That supports teaching verification, rather than simply encouraging adoption.
Academy OS should help people recognise assumptions, check sources, challenge answers, record uncertainties and identify when specialist help is needed. Ownership alone is insufficient if someone cannot assess what their system produces.
5. The findings deserve respect, with limits.
This is a substantial survey, but it concerns Vanguard investors. The sample was 60% baby boomers or older and 59% advised clients; headline estimates were weighted for age and advised status. It is a snapshot of attitudes in June 2026, rather than evidence of long-term behaviour or a representative picture of everyone needing financial help.
The paper also comes from an institution that provides investment and advice services. That does not invalidate the findings, but its recommendations understandably interpret demand for human support through an adviser-centred lens.
My reading is that the Academy should use this as evidence for its direction, not as proof of its commercial model. Your practical test remains whether people gain capability, what triggers them to seek support, what they will pay, and how much human time delivery requires.
The memorable distinction is:
Demand for human help is not necessarily demand for human control.
Comments