The AI-Informed Client

Introduction
As CEO of Agora, I spend a growing share of my time in conversations about AI and advice. Most advisors I talk to aren’t afraid of technology. What they’re afraid of is being caught flat-footed in front of a client.

Many are incorporating AI tools at their own pace, in their own way — which is natural. But a structural flaw of the independent channel is that we tend to work in silos. I have the advantage of talking across dealer groups and across the country. What I see is a lot of parallel experimentation, and not enough shared learning.

I’d like to change that. The paper below is an attempt to open the conversation — and to give independents a framework for getting ahead of AI rather than being pulled along by it.

The practice is arriving faster than the literature. There is not yet a deep body of research focused specifically on the AI-informed client in wealth advice, even as advisors are beginning to encounter this behaviour in real client relationships. Advisors are meeting this shift before the industry has produced a widely shared framework for how to interpret it, discuss it, and respond to it with confidence.

The medical profession offers an instructive parallel. When patients began arriving at appointments with printouts from WebMD and then, later, far more sophisticated AI-generated symptom analyses, physicians were similarly caught without a common framework. The “Dr. Google” phenomenon — a term that entered clinical literature in the mid-2000s — forced the profession to confront a question it had not anticipated: what is the role of clinical expertise when the patient arrives already informed? The financial advice profession is now encountering the same question, and the window to build shared norms proactively — before the behaviour becomes universal — is open, but it will not stay open indefinitely.

This paper is intended to open that conversation. Its purpose is to name the AI-informed client as an emerging and important development in advisory practice, to offer a structured way of thinking about it, and to encourage advisors to reflect on what they are already seeing in the field.

It is also part of a broader effort to gather data directly from advisors. Alongside this paper, Agora will collect survey responses on how advisors are encountering AI-informed clients, where they feel confident, where they want more support, and how these interactions are shaping client relationships. Those responses will inform a follow-up publication that shares the findings back with the advisor community.

The starting point is simple and often overlooked: a client who uses AI to prepare for a meeting is a highly engaged client. They have framed a financial question, invested their own time, and arrived with a point of view. That behaviour is a signal worth reading carefully because it creates the conditions for a more substantive advisory conversation.

At Agora, we see AI-informed clients as the leading edge of a broader shift in client behaviour. Canadians are becoming more comfortable using AI as a research tool, yet still want human judgment, context, and accountability when decisions matter. That combination creates an important opening for advisors who can meet these moments with clarity and professionalism.

This whitepaper offers one framework for doing that. It is built around four pillars:
• AI-informed clients are signalling engagement, and that engagement is an asset advisors can work with productively.
• How an advisor responds in the first AI moment with a client becomes a trust inflection point in the relationship.
• AI changes what expertise looks like, but it does not diminish its value — it redirects it.
• AI literacy is becoming a professional competency, not a personal preference.

“Wealth isn’t simple. Advice matters.”

Pillar One: The Engaged Client Thesis
AI familiarity is growing — unevenly, and consequentially
The AI-informed client is not a uniform phenomenon. It is concentrated among specific demographics, and understanding that distribution matters for how advisors read the behaviour when they encounter it.

The 2026 TD AI Insights Report, based on an Ipsos survey of 2,501 Canadians, puts hard numbers to what many advisors are beginning to feel in practice. AI familiarity varies sharply by generation: 87% of Gen Z and 77% of Millennials describe themselves as familiar with AI, compared with 61% of Gen X, 40% of Boomers, and 39% of Boomers+.

The implication for advisors is not abstract. An advisor with a mixed-age book of clients is effectively operating across two very different client realities simultaneously. A 34-year-old client arriving with an AI-generated retirement projection and a 67-year-old client who has never used an AI tool are both in the same advisor’s calendar — and they require meaningfully different conversational approaches.

87% of Gen Z Canadians describe themselves as familiar with AI (TD AI Insights Report, 2026)

39% of Boomers+ say the same — a 48-point gap that lands squarely in an advisor’s client book

This is not simply a generational curiosity. The cohort with the lowest AI familiarity — Boomers and Boomers+ — holds the majority of investable assets in Canada. The cohort with the highest familiarity — Gen Z and Millennials — represents the next decade of wealth accumulation and intergenerational transfer. Advisors are navigating both simultaneously, and the AI-informed client dynamic will intensify as the familiarity gap closes over time.

What this means in practice is that the AI-informed client today is predominantly a younger, higher-engagement client who has already decided to invest time in preparing for a financial conversation. That is a fundamentally positive signal. The advisor who responds to it well — who treats the client’s AI output as a starting point rather than an intrusion — builds credibility with the cohort that will define the next generation of advisory relationships.

Engagement as a professional asset

The parallel with medicine is instructive here. Research on physician-patient communication has consistently shown that patients who arrive with information — even imperfect information — are more likely to follow through on treatment plans, ask better questions, and report higher satisfaction with the clinical encounter. The challenge was never the informed patient. The challenge was the physician who responded to patient research with dismissal rather than engagement.

Financial advisors are at a similar inflection point. The client who arrives with an AI-generated analysis is demonstrating initiative. They have a hypothesis. They want to be taken seriously. How the advisor responds in that moment — whether they engage the analysis or deflect it — sends a signal about the kind of professional relationship on offer.

Advisors who can say, with genuine confidence, “Let’s look at what you found and build on it,” are not conceding authority to an algorithm. They are demonstrating the kind of intellectual engagement that differentiates professional advice from generic guidance.

Pillar Two: The Trust Inflection Point
Where Canadians still want humans

AI familiarity and AI trust are not the same thing. The TD data makes this distinction with unusual clarity: even among the most AI-familiar cohorts, Canadians draw a consistent line between tasks where AI is welcome and decisions where it is not.

More than half of Canadians express a preference for human-only interactions when it comes to: financial planning advice (55%), assessing approval for a financial product (55%), and planning for retirement (53%). Even budgeting — a relatively routine financial task — sees 44% preferring human guidance.
Only 1 in 3 Canadians say they would rather trust AI over a human — even their parents — for financial advice. The instinct toward human accountability on high-stakes decisions is not a lag in technology adoption. It is a considered preference that persists even among people who use AI regularly.

1 in 3 Canadians would trust AI over a human for financial advice — meaning 2 in 3 would not (TD AI Insights Report, 2026)

This maps almost exactly onto the advisor’s daily experience. The AI-informed client is not arriving to replace the advisory relationship. They are arriving better prepared for it. The TD data confirms what advisors are beginning to feel: clients are using AI as a research tool, not as a substitute for professional judgment.

What drives distrust — and what builds it back

The TD survey identified three primary drivers of AI distrust: inaccurate information, privacy and security risk, and lack of accountability.
Each of these is, notably, an area where a skilled human advisor provides direct and visible value. The advisor who corrects an AI-generated misunderstanding is demonstrating accuracy. The advisor who explains how client data is handled is addressing privacy. The advisor who takes explicit responsibility for their recommendations is providing accountability. The three things that drive clients away from AI are precisely the things a good advisory relationship delivers.

When asked what companies could do to increase trust in their use of AI, respondents pointed to: offering human oversight or ways to intervene, taking responsibility if something goes wrong, and protecting data and privacy. The trust architecture clients are asking for is not “no AI” — it is AI with a human in the loop who is accountable for outcomes.

The things that drive clients away from AI are precisely the things a good advisory relationship delivers: accuracy, privacy, and accountability.

This is the trust inflection point. The advisor who positions themselves as the accountable human in an AI-assisted process — who can say, “I use AI to be better prepared for this conversation, and I stand behind the advice I give you” — is offering exactly what the TD data says clients want.

The Dr. Google parallel

Physician-patient communication research offers a useful framework for thinking about this moment. When the “Dr. Google” phenomenon emerged, the instinct of many clinicians was to dismiss or correct patient-sourced information quickly, in order to reassert clinical authority. The research showed this was counterproductive. Patients who felt dismissed were less likely to disclose their self-research in future encounters, more likely to act on it without guidance, and less likely to follow clinical recommendations.
The more effective clinical response was acknowledgement first: “You’ve done some research on this — let’s look at what you found.” This approach, documented in teach-back and shared decision-making literature, preserved the clinical relationship while creating space to correct errors and add professional context. The clinician did not become less authoritative by engaging the patient’s research. They became more trusted.

The financial advice profession is at the same crossroads. Advisors who can engage AI-generated client research — acknowledging its strengths, contextualizing its limitations, and building on it professionally — will emerge from this transition with stronger client relationships than those who do not.

Pillar Three: Expertise Redefined
What AI changes about the advisory role

The most common anxiety among advisors encountering AI-informed clients is that expertise is being commoditized. If a client can generate a retirement projection, a portfolio allocation, or a tax scenario in sixty seconds with a language model, what is the professional value of an advisor who provides the same output?

This anxiety, while understandable, misidentifies where advisory expertise actually lives. The advisor’s value was never primarily in producing outputs. It was in knowing which outputs to produce, how to interpret them in the context of a specific life, and how to translate them into decisions a client could understand and act on with confidence.

Research on AI financial advice quality is instructive here. Studies from MIT and Stanford have found that AI-generated financial guidance can be technically accurate and, in some cases, comparable to professional advice on standardized financial questions. But the same research found that the quality of AI output is highly sensitive to the quality of the prompts that generate it — with higher financial literacy in prompts modelled to produce roughly 5% better investment outcomes.

The implication is significant: the clients best positioned to use AI well are those who already have financial sophistication. The clients least equipped to use AI well — those who most need professional guidance — are also least likely to recognize the limitations of the output they receive. The advisor’s role as a quality filter is not diminished by AI. It is made more consequential.
There is also a category of advisory value that AI cannot replicate at all: the judgment that comes from knowing a specific client’s circumstances, relationships, risk tolerance, and life context over time. AI can model a retirement scenario. It cannot know that the client’s health has changed, that their son is getting divorced, that they’ve been lying awake at night about market volatility. That knowledge — accumulated through trust over years — is the foundation of advice that actually changes outcomes.

The sense-maker, not the information holder

The medical parallel is again useful. The physician’s role did not disappear when patients gained access to medical databases. It shifted. The physician became less valuable as a source of information — a function that databases could partially replicate — and more valuable as a sense-maker: someone who could take a patient’s collection of symptoms, history, and internet research and turn it into a coherent clinical picture and a plan.

Financial advisors are undergoing the same role evolution. The advisor who embraces this shift — who positions themselves explicitly as the professional who makes sense of a client’s AI research rather than competing with it — is operating from a position of genuine strength. The advisor who tries to compete with AI on output speed and breadth is not.

This reframing has practical implications for how advisors present their value. The conversation is no longer “I will tell you what to do with your money.” It is “I will help you understand what you’re looking at, put it in the context of your specific situation, and make sure the decisions you make are ones you can stand behind.” That is a more accurate description of what good advisors have always done. AI has simply made it easier to articulate.

Pillar Four: AI Literacy as Professional Competency

From personal preference to professional standard

The question of whether a financial advisor should develop AI literacy is no longer a lifestyle choice. It is becoming a professional competency in the same way that digital literacy became a professional competency in the 2000s. Advisors who were slow to adopt digital communication tools did not disappear — but they operated at a structural disadvantage in client acquisition, relationship management, and practice efficiency that compounded over time.

The dynamic with AI literacy is similar, but the pace is faster and the gap between early adopters and laggards will close less forgivingly. The advisor who understands how AI tools work, what they are good at, what they reliably get wrong, and how to use them to prepare better client conversations is operating from a fundamentally different capability base than the advisor who does not.

The TD data adds a dimension that is easy to overlook. The 58% of Canadians comfortable with financial institutions using AI behind the scenes for fraud detection, the 57% comfortable with AI-assisted spending analysis, the 55% comfortable with AI product recommendations — these numbers reflect a population that is already accustomed to AI in their financial lives, even if they don’t always name it as such.

Clients are not arriving from a world without AI. They are arriving from a world where AI is already embedded in their banking apps, their credit decisions, and their investment platforms. The advisor who can speak fluently about how AI is and is not being used in the advisory relationship is meeting clients where they already are.

What AI literacy looks like in practice

AI literacy for financial advisors is not about becoming a technologist. It is about developing a working understanding of three things: what AI tools can and cannot do reliably in a financial context; how to recognize the limitations of AI-generated analysis when clients present it; and how to use AI tools selectively to be more prepared, more efficient, and more present in client conversations.

The advisor who can look at a client’s ChatGPT-generated portfolio analysis and say, with genuine authority, “This projection makes a standard assumption about returns that doesn’t account for your tax situation or your income variability — let me show you what changes when we adjust for that” is not threatened by the client’s AI use. They are demonstrating the professional judgment that AI cannot replicate.

In that sense, AI literacy is not separate from trust-building. It is part of it. The advisor who can engage confidently with client-generated AI output is not simply demonstrating technical competence. They are demonstrating readiness for the way advisory relationships are evolving.

Where This Leads

The AI-informed client is not a fringe case. It is an early signal of a broader shift in how clients prepare, question, and participate in the advisory relationship. The TD data confirms that this shift is already underway: Canadians are growing more familiar with AI, more comfortable with it in routine financial contexts, and no less insistent on human judgment for the decisions that matter most.

That combination — AI-prepared clients who still want human advisors — is not a contradiction. It is an opportunity. The clients arriving at advisory meetings with AI-generated analysis are not trying to bypass their advisors. They are trying to arrive ready for a better conversation. The advisors who meet them there will build the strongest relationships of their careers.
Medicine offers an instructive precedent. The profession’s encounter with the Dr. Google patient unfolded over more than a decade, and the norms that eventually emerged — around how to acknowledge patient research, how to correct it without dismissing it, how to reframe the clinical role in a higher-information environment — were built largely through trial and error. The financial advice profession has an opportunity to move more deliberately. The behaviour is still concentrated among early adopters. The pressure points are visible but not yet universal. The window for building shared frameworks, training approaches, and professional language before this becomes the default client experience is open.

This paper is meant to advance that conversation. It offers one way to think about a new client dynamic that many advisors can already feel, even if they do not yet have a settled language for it. The next step is to complement this framework with direct input from advisors themselves.

Agora will be collecting survey responses as a follow-on to this paper in order to better understand where advisors are encountering AI-informed clients, how they are responding, and what patterns are beginning to emerge across the profession. Those findings will help sharpen the framework, identify the most important pressure points, and contribute to something the financial advice profession currently lacks and demonstrably needs: a shared, evidence-based body of practice for navigating the age of the AI-informed client.

About Agora
Agora Wealth Corp. is a Canadian registered custodian and wealth management platform serving the mass affluent market. Agora works with a national network of dealer groups and financial advisors, combining custodian infrastructure, a TAMP layer, and proprietary advisor technology to support better client outcomes. “Wealth isn’t simple. Advice matters.”