Financial advisors often debate how and whether they should use generative AI tools like ChatGPT, Claude and Gemini, especially when it comes to managing clients. But advisors are not the only ones who can access the technology. Clients are already using AI to generate their own investment ideas and inform financial decisions.
Nearly 60% of U.S. affluent investors use AI for financial and investment tasks, according to a June survey by HSBC.
For advisors already worried about AI replacing them, the prospect of a client telling them they have consulted AI about financial issues can be disconcerting. That’s especially true if AI said something that contradicted the advisor.
But experts say advisors should not react defensively.
Advisors should respond “by first taking a deep breath and reminding themselves that this is not a threat or an attack on their credibility,” said Brendan Frazier, Host of The Human Side of Money podcast. “And from there, keeping a mindset of humble curiosity. Reaffirm that this is a great question. And then use it as an opportunity to dive deep and understand why the advice conflicts.”
Just as patients have used WebMD and other online health care sources before seeing their doctors, advisors should see this as an opportunity to further engage and collaborate with clients.
“Tell them that it’s identical to how you would use it for anything medical-related,” Frazier said. “You can use it to become more educated and informed faster. But, you don’t ultimately want to act on that information because it’s giving you the best possible answer based on the information it has. But you ultimately still want to talk with a doctor before making any big decisions.”
Indeed, the HSBC survey said only 7% of U.S. investors cited AI as the most influential factor in their last investment decision. And more than one-third (38%) of investors said they prefer a hybrid approach (AI and advisors) to their decision-making.
“AI is now part of every client’s toolkit, and advisors shouldn’t resist it,” said Anshuman Mehta, Chief Product Officer of Domify. “Treat a conflicting AI recommendation as a discovery opportunity. Why did the client seek a second opinion? Have their goals, risk tolerance or life circumstances shifted, perhaps disclosed to a chatbot but not their advisor? Ultimately, advisor and client want the same outcome: what’s best for the client.”
AI Vs. Advisor
Advisors should be prepared for clients who bring them AI advice that doesn’t necessarily mesh with the advisor’s opinions. In fact, the AI might be correct.
“Don’t be combative,” said Justin Whitehead, Co-Founder and CEO of Pebble Finance. “Try to understand why the AI reached a different conclusion. Does it lack the full picture? Or does the advisor? AI is only as good as the information and assumptions behind it. And advisors need to be open to the possibility that the AI identified something they missed. People make mistakes, too.”
Said Dan Daum, Co-Founder and CEO of WealthStream: “Start by asking what the client found compelling. Then walk through how the two recommendations differ and what each assumes about their situation. If the AI caught something you missed, say so. Clients need to know you’re willing to reconsider. That can build more trust than insisting you were right the first time.”
Advisors Still Driving
Experts say advisors should openly welcome the AI-generated information clients bring to their attention. But advisors should demonstrate the value they bring to the relationship by providing the proper context and expertise to help clients make sound financial decisions.
“I’d encourage clients to use AI,” Whitehead said. “An informed, curious client is a good thing. But there’s a difference between getting an answer and making a financial decision. Advisors can help clients understand the tradeoffs, challenge assumptions, and connect the analysis to their broader financial picture.”
“AI doesn’t change why advisors exist: to offer an objective, dispassionate view of client goals, insulated from market emotion,” Mehta said. He said advisors can use AI to bring together account data, trade history, investment policy statement objectives, emails, meeting notes and CRM entries to build a holistic client view and enrich conversations.
Frazier said advisors should position themselves to clients not as mere providers of advice but as a “thinking partner and collaborator.”
“It’s important to first recognize that a client looking for this information is an engaged client, which is ultimately a good thing,” Frazier said. “And the worst thing that you can do is try to bash AI or convince them that they shouldn’t even be using it, because people trust AI (whether they should or not). And then use their questions as an opportunity to demonstrate where AI has strengths and weaknesses and educate them on the best ways to use AI and then the best times to rely on you as their thinking partner and collaborator.”
Reviewing Chats
Our experts were split on whether advisors should ask clients to provide them with copies of their conversations with chatbots. Some believe that the conversation will provide advisors with clues as to why the client consulted AI in the first place.
“The prompt often reveals what a client confides to a chatbot but not their advisor, such as evolving life circumstances, liquidity needs or shifting return expectations,” Mehta said. “It can also surface data outside the advisor’s view, like held-away assets, rounding out the advisor’s view of client assets.”
Asked whether advisors should request the chat, Daum said, “Yes, if they’re comfortable sharing it. Seeing the conversation helps you understand what the client asked, what information they provided, and how the answer developed. Sometimes the questions tell you more than the answer. They can reveal a worry or priority that hasn’t come up in your meetings.”
But Frazier generally advised against asking to see the prompt or chat. He made an exception when specific inputs could directly affect the advice and incorrect information could produce a harmful recommendation.
“As a general rule, no, because it signals that you don’t trust the client’s ability to ask/prompt the right way,” Frazier said. “Similar to when your kids say they finished their homework, and your first response is ‘Let me see!’ It subtly signals to them that you don’t trust them.”
“But if the actual specifics of the prompt have a direct impact on the quality of the advice to the point that one wrong piece of information could result in a harmful recommendation, then you can ask for it in a friendly way: ‘Would you mind showing/reading me what you put in the prompt so I can verify that we are all working off of the same information? This is one of those situations where one number being off could change the entire recommendation,’” Frazier said.
“The prompt or chat history won’t capture all the information that influenced the AI’s recommendation,” Whitehead said. “I’d focus on understanding the recommendation itself, the assumptions behind it, and the evidence supporting it. The advisor’s job isn’t to audit the chatbot. It’s to evaluate the advice.”
AI Pitfalls
Advisors should take the opportunity to educate clients about AI and what it can or can’t do. Clients should closely examine AI’s recommendations and why it made them.
“Encourage clients to ask follow-up questions,” Daum said. “What are the downsides? What information is missing? What would change this recommendation? They should check key claims, avoid sharing sensitive information, and talk through major decisions with their advisor. An answer can sound convincing and still leave out something that matters.”
Said Whitehead: “AI can sometimes gaslight you, confidently reinforcing assumptions that may be wrong. One trick I use is asking AI to challenge its own conclusions and make the strongest argument against its previous answer. Financial decisions rarely have one objectively correct answer. Understanding the competing arguments is often more valuable.”
“A simple demonstration is to run the same scenario through AI several times and compare the outputs, which are often materially different each time,” Mehta said.
Pricing The Advisor
If AI can perform increasingly sophisticated tasks, then clients might naturally question the fees they pay advisors. So advisors should remind clients what they bring to the table.
Frazier said that as financial knowledge becomes more accessible, advisors’ value increasingly lies in providing wisdom, guidance and perspective and helping clients apply information to their particular circumstances.
“The advisor’s value shifts towards being a thinking partner and collaborator helping clients separate the signal from the noise in order to determine the best decision for their specific situation,” Frazier said.
Said Daum: “Clients will expect more for what they pay. The advisor needs to show how they’re spotting issues, helping clients weigh difficult choices, and making sure things get done. I think that creates pressure to provide more comprehensive advice for the same fee — advisors who use AI to deliver that have a stronger case for what they charge.”
“As AI makes sophisticated financial analysis increasingly accessible, the value shifts from providing answers to knowing which questions to ask,” Whitehead said. “That’s harder than it sounds. Understanding a client’s circumstances, identifying what matters, and exercising judgment become more valuable. Advisors charging primarily for information will face pricing pressure. Those delivering better decisions may not.”
Larry Roth is CEO of Wealth Solutions Report and Founder and Managing Partner of Ascentix Partners.
