With 86% of Gen Z and millennial investors now using AI regularly, advisors explain what the shift means for advice — and where AI dangerously falls short
The numbers from Bank of America’s 2026 Study of Wealthy Americans are striking: 86% of Gen Z and millennial investors now regularly use artificial intelligence, 86% are comfortable incorporating it into their financial lives, and 87% believe it is important that their financial institutions use advanced technology. Yet the same study found that 71% worry AI can provide incorrect information and 75% have concerns about privacy and data security. The result is a generation of investors who are simultaneously the most technologically empowered and the most skeptical cohort to sit across from a financial advisor — and who are reshaping what the advice conversation has to be.
Wealth managers who work directly with AI-literate clients say the implications are more nuanced, and more urgent, than most of the industry’s AI commentary has acknowledged.
AI can educate, but it cannot read the room
Megan Miller, senior wealth advisor and managing director at MAI Capital Management, says AI can be a genuinely useful supplemental tool for clients in the early stages of building wealth. The problem, in her experience, is not the information AI provides — it is everything it cannot perceive.
“The biggest shortfall we see is the current inability for AI to understand and tailor advice to individuals’ and couples’ needs beyond the numbers,” Miller said. “Clients want a connection and a relationship that AI cannot provide. Wealth management is about understanding people and their needs. AI only scratches the surface on everything we cover with our clients and their families.”
The most common mistakes Miller has observed from AI-reliant younger investors fall into two categories. The first is focusing purely on quantifiable inputs while ignoring factors that do not fit neatly into a prompt — potential inheritance, differing income trajectories, and the psychological histories that shape how clients respond to financial stress. The second is structural: when one member of a couple runs a financial scenario through an AI tool, the output reflects only that person’s inputs and perspective, systematically excluding the other partner’s goals, concerns, and risk tolerance.
“Ultimately, they want to know our opinion, as we have the long-term relationship and can compile all of their financial needs and considerations that AI wouldn’t even know to ask about. AI can’t yet read nonverbal cues, body language changes and unsaid dynamics between a couple or a family, which are critical for advisors to deliver impactful and personalized advice,” Miller said.
General-purpose AI makes dangerous errors on investment math
Dr. Toby Wade, co-founder and CEO of DeepVest, brings a data-specific warning that advisors need to hear clearly: the AI tools most widely used by the investing public are fundamentally unsuited to financial calculations.
General-purpose large language models like ChatGPT are trained on internet-scale text data, not curated financial datasets, and they operate probabilistically — predicting the next most likely token in a sequence rather than computing deterministic answers. That architecture works well for writing, summarizing, and brainstorming. It fails, sometimes catastrophically, when applied to tax-loss harvesting calculations, cost-basis analysis, or risk-adjusted return metrics like Sharpe ratios.
“In our research, comparing general-purpose AI models against institutional-grade investment analysis, we documented error rates exceeding 85% on investment analytics tasks. Advisors are right to be skeptical — not of AI broadly, but of applying consumer AI tools to domains that require precision and high-quality data,” Wade said.
Wade’s broader observation about the advisor’s role, however, is optimistic. Younger, digitally native investors arrive at meetings having already run their own research through AI tools, podcasts, and social media. The question they bring has fundamentally changed — from “what should I do?” to “why is this right for me?” Wade draws a parallel with the shift in medicine, where patients now arrive at appointments having already reviewed their own lab results through AI tools and come ready to question their doctor’s interpretation. The same dynamic, he argues, has arrived in wealth management and represents a net improvement in the depth of client conversation.
AI is collapsing the middle — judgment is what remains
Rick Nott, managing director at Angeles Wealth Management, offers what may be the most structural framing of what AI actually changes in the advice business.
In Nott’s model, the value chain of financial advice runs from raw data at the bottom to decisions and action at the top, with a wide band of intermediate work — gathering, formatting, reconciling, summarizing — that has historically consumed a disproportionate share of advisor time without adding proportionate client value. AI is compressing that middle layer rapidly.
“For decades, service professionals spent an absurd amount of time on everything in between: gathering, formatting, reconciling, summarizing. Work that was necessary, but not always additive to our clients. AI is collapsing that middle. Years from now, I think we’ll look back on this the way we look back on paper and fax machines before email. The part that always mattered is still the part that matters now: the judgment and action at the top,” Nott said.
Nott is direct about the limits of AI in high-stakes financial decisions. The tools are, as he puts it, famously eager to agree with the user — a characteristic that makes them particularly dangerous when a client is at their most emotionally activated. Better math, he argues, has never been the primary bottleneck to better financial outcomes. The real work of financial advising is behavioral: helping clients build the mental models and habits that allow them to act against their instincts at precisely the moments when instinct is most likely to lead them wrong.
“All money decisions are emotional decisions, directly or indirectly,” Nott said. “A smarter-sounding tool doesn’t fix that. Financial success is about building a system of habits and mental models — a family financial operating system — that lets you do what you’re wired not to do: get aggressive when the world looks scary, get conservative when everyone else is greedy, and reconcile your own money beliefs against your family mission.”
Taken together, the three perspectives point toward a consistent conclusion: AI is not displacing financial advisors, but it is forcing a rapid and permanent upgrade in what the advisor role requires. The clients who are most enthusiastic about AI — younger, digitally native, already arriving with pre-formed conclusions — are also the ones who most acutely expose what AI cannot do. The advisors best positioned to serve them are those who can move fluently between technological fluency and the irreducibly human work of judgment, relationship, and behavioral coaching.
