As artificial intelligence becomes a bigger part of client conversations, advisors need to be aware of the capabilities of chatbots such as ChatGPT and Google AI, whether that’s good or bad.
New research shines a light on these tools that are becoming mainstream for consumers and slowly gaining traction in the wealth management industry to find out if clients and even advisors can rely on the technology in a key area of advisory practices.
The study, published today (9/23), reveals that AI tools can be unreliable when used for estate planning, leaving users at risk of inaccurate information in a field where accuracy and nuance are non-negotiable. It highlights how human advisors remain key to estate planning advice.
The EncorEstate Plans study assessed four widely used platforms (ChatGPT, Claude, Perplexity, and Google AI Mode) by asking each of them 46 estate planning questions drawn from the most frequent client inquiries that the firm’s planning team has fielded over the past 45 years.
Answers were scored on a traditional A–F scale and the results are sobering. The top performer, Claude, achieved 69% of answers graded A or B, while Perplexity followed closely at 63% A or B responses.
Probably the best-known AI chatbot, ChatGPT, performed significantly worse than the two top-rated options tested, with nearly half of its answers earning only a D or F.
Meanwhile, Google AI Mode performed worst of all with 61% of responses failing, before it stopped after question 19, leaving 28 questions unanswered.
That means that a client who relies on an AI chatbot for guidance could end up with a trust riddled with errors or omissions and EncorEstate CEO Matt Morris says that his team has seen clients attempt to finalize AI-generated documents containing “nonsense POA powers” or questionable clauses that would never hold up in practice.
Even when chatbots do provide complete answers, they often fail to account for client-specific context. That gap is where professional oversight is essential.
The takeaway from the study is not that advisors should ignore AI, but rather that they should set boundaries for its use. Tools like Claude and Perplexity may offer reasonable starting points for basic education, but they cannot replace professional review.
For advisors, the opportunity lies in reinforcing their value and impressing on clients that estate planning is not just about documents, it’s about judgment, context, and guiding families through complex choices.
Chicago-based Mesirow Fiduciary Solutions adds flexPATH's plan-level outsourced fiduciary book, boosting its retirement market reach to $164 billion.
Also, Raymond James's employee advisor channel adds breakaways from Wells Fargo and Stifel, while UBS welcomes an ex-Morgan Stanley duo in Indiana.
It promised guaranteed principal and 10% monthly returns. The SEC says it invested nothing.
Ten years after entering the fee-based annuity market, MassMutual Ascend says nearly half its lifetime sales came in the past two years alone – but barriers remain among fee-only advisors.
Amplify, WealthReach, Zeplyn and Zocks have unveiled partnerships aimed at automating portfolio management, content creation, account opening, and client communication.
Northern Trust’s Ken Lassner shows advisors how to convert volatility into after-tax portfolio gains
Dan Biagini of American Equity says the steady decline of pensions, longer lifespans and a reset in interest rates are rewriting how advisors build retirement income
