I was recently talking with a friend who has been wanting to do Financial Planning for several years now but her spouse isn’t on board. Her spouse is smart and resourceful, so with the advent of AI they have been consulting a chatbot for financial planning advice rather than a human planner. She confided that she feels nervous about what she doesn’t know and is even more nervous about relying on artificial intelligence to help them make such an important decision as to when she can retire. In her words, it seems, “too important to put in the hands of AI”. She asked me my thoughts, and this is what I told her.
Outputs are only as good as the inputs
When an untrained individual is entering data into a chatbot their answers will only include what they think they need to include, leaving room for information to be missed. This can happen with a human advisor, too, but it has a greater chance of being caught than it would with a computer. For instance, we recently delivered a Financial Plan to a couple whom we had been gathering data from for close to 2 months. We asked many probing questions and requested the usual data- past tax returns, account and policy statements, paystubs, a list of expenses, and more. But, as is often the case, several important details came up in our meeting that they had forgotten to disclose! One of their mom’s had died a few years ago and they are still trying to settle the estate which will provide an inheritance; One of them has a sibling who is an addict and is irresponsible with money so they want to make sure he can’t inherit anything from them upon their death; One has a long-term health issue that could cause an eventual long-term care situation potentially costing tens of thousands of dollars, if not more; and one of them has an old life insurance policy that’s just been wasting away without premium payments. All of these details came out in our organic conversation, and all of it is extremely important when planning for the long-run. If you don’t put details like these into the chatbot, you’ll be missing crucial pieces of the puzzle making your financial planning results misinformed.
It’s easy to misinterpret complex instructions
I’ve met several people with horror stories that began with misunderstanding technical instructions given to them by an industry rep. We all like to think it won’t happen to us, but when you’re dealing with an industry that is completely foreign to you, it is all too easy. Here’s an example of such a misunderstanding: Jenny recently got divorced and it was decided that she needs to give ½ of her IRA to her ex-spouse. Jenny is only aware of one way to accomplish this task, which is withdrawing the funds from the account and paying the proceeds outright to her ex. So she completes a withdrawal form withdrawing ½ the account balance plus more to cover the taxes so the withdrawal nets out ½ the account value for her ex. Though this accomplishes the end goal of splitting the asset, it isn’t ideal as now the withdrawal will be taxable to her rather than her ex, which means it will add to her taxable income and possibly boost her up into a higher tax bracket. It also means she won’t be left with half the account value, she’ll be left with slightly less because she needs to withdraw more money to cover the taxes so that her ex can receive the full ½ of the account value. What’s more is, her ex will now miss out on receiving the built-in tax-deferral benefits of an IRA because they will be getting cash. Instead of this, she could have handled it according to industry standards where she would provide a copy of the divorce decree to the IRA custodian and they would in turn transfer ownership to her ex behind the scenes. Doing it this way preserves the tax-deferral of the account, saves Jenny from paying taxes or any potential penalties, and ensures they both get ½ of the account value in the end. This example clearly shows why you should not rely solely on impersonal instructions. If this isn’t your area of expertise you can’t assume you will accurately know how to implement the information you receive, regardless of how smart you are, and it can end with irreversible mistakes.
You don’t always know what to ask
Many years ago, when I first started in the industry, I was working jointly on a case with a more seasoned advisor. When reviewing a client’s estate plan we found that his plan was to fully liquidate his retirement accounts now, while he was still alive, and move the proceeds as well as other assets into an irrevocable trust, all with the intention of avoiding estate taxes. What he didn’t realize was that by doing this he was not only going to pay an exceptional amount of income taxes now but he was losing the retirement account’s tax deferral benefits and his heirs would lose the IRA stretch capabilities as well as the eventual step-up in basis on his properties. Did he accomplish his main objective of not paying estate taxes? Yes. But was it the best solution for him? Probably not. This case perfectly exemplifies how we don’t always know what to ask which can lead to bigger problems than what we started with.
You have no accountability or recourse
AI cannot provide the same level of accountability as a human can. Our clients all get homework lists that we review and debrief together in our monthly meetings to make sure there is progression. And if something goes wrong, we can help fix it. With AI there is no recourse, you’re just stuck. It can’t provide help to fix the issue nor can you take it to the courts to have them fix it. AI just simply can’t replace a human in these two situations.
It doesn’t know everything
Last summer we had an intern join our team. When we were working on a tax-related project he consulted AI for the then-current tax brackets. Later, when I was reviewing his work, I noticed the tax brackets he was using were way off. I was shocked when I learned he had gotten them from AI because I figured something as unwavering and widely-known as IRS tax brackets would be easy enough to get right. Unfortunately, every bracket of the 7 different brackets were listed completely wrong. In fact, it was baffling just how wrong they were! What’s more is, the inaccuracy has remained an ongoing problem and still exists today!
Another example of inaccurate reporting came more recently. I like to keep a pulse on the Socially Responsible Investment (SRI)/Environment Social Governance (ESG) investment options available in the marketplace as it’s one of my niche offerings. A few months ago I asked AI to give me a list of all the SRI/ESG mutual funds and ETFs available in the marketplace. I expected the program to generate an enormous list that would include multiple share classes with all the various fund minimums and take forever to load, but, quite the opposite happened. Less than 15 investments were listed and what were missing were the oldest, most established SRI/ESG money managers! I tried again yesterday only to have the same thing happen, where only a few obscure options were listed, making me less convinced that AI knows as much as we think it knows!
What if the information is manipulated?
My last point is somewhat subjective and admittedly skeptical. Many people have seen examples of AI chatbots being influenced, whether intentionally or unintentionally, to produce biased or inaccurate responses. When we know that these systems can be shaped by the priorities, perspectives, or oversights of their creators, it raises the question if we should fully trust the information AI provides? And if that’s the case, how useful is the tool if its accuracy isn’t reliable?
The last word on the subject, provided by MS CoPilot
After writing this blog post I thought I would check with AI to see what it has to say about the subject. According to Microsoft’s Copilot on 5/28/2026, the top reasons why you should not use AI as your main Financial Planner are as follows:
1. AI can’t fully understand your personal context. (Family dynamics, emotional risk tolerance, personal values, career uncertainty, etc.)
2. AI can sound confident while being wrong. It can misinterpret your situation, overgeneralize things with incomplete data, produce outdated or oversimplified guidance, or miss regulatory or tax nuances.
3. There is no Fiduciary Responsibility. No accountability, no recourse, no liability.
4. Weakness with tax, legal, and regulatory complexity. Financial planning intersects with tax law, estate planning, retirement rules, insurance regulations- all of which change constantly and differ from state to state.
5. No emotional coaching during market stress. There’s no talking you out of panic selling, or helping you remain disciplined, or keeping your long-term goals in focus.
6. Privacy and data-handling concerns. Relying solely on AI means more data is stored digitally, there’s more risk if systems are breached, and there is less control over how your information is used.
7. AI can’t coordinate with other professionals. A real financial planner collaborates with your accountant, attorney, insurance agent, and HR/benefits team.
8. AI can’t anticipate life events the way humans can. Humans ask probing questions like, “Are you going to have kids?”, “Is your job stable?”, and “are you caring for aging parents?”
9. Overreliance creates a false sense of security. It may encourage you to: stop learning about your own finances, miss red flags, fail to question assumptions, and accept recommendations without scrutiny.
There’s no doubt that AI is an impressive tool. I’ve use it and can definitely appreciate it, but it’s still just a tool. Financial planning is deeply personal, filled with nuance, emotion, and long‑term consequences that extend far beyond numbers and mathematical calculations. A chatbot can process data, but it cannot catch the details you forget to mention, anticipate the life events you don’t think to bring up, or guide you through the human side of money decisions. It cannot coordinate with your attorney, understand your family dynamics, or take responsibility when something goes wrong. For something as important as your financial future, AI can support the process, but I agree with my friend, financial planning is far too important to leave in the hands of AI.