Intuit, the company behind Mint and TurboTax, recently said it was trying to make financial planning more accessible by incorporating generative AI into its software.
“Not everyone is born with the financial knowledge needed to know how to research and ask the right questions,” said Nhung Ho, vice-president of AI with Intuit, whose team has spearheaded the company’s new generative AI operating system, GenOS.
“What we’re trying to do is meet people where they are and allow them to ask questions in their own words, which the program can use to match them up with the best advice and ways forward.”
Though complex generative AI systems are a novelty within the financial planning world, AI has been a significant part of personal finance services for nearly a decade, said Alberto Rossi, a professor of finance and robo-advising at Georgetown University in Washington, D.C.
He pointed to financial aggregators like Intuit’s Mint, which tracks customers’ bank accounts to help manage their budgets and cash flow, that emerged in the early 2010s as predecessors to today’s GenOS-esque programs.
“As technology changed, companies started realizing that a lot of the market was cut out of financial advice,” said Rossi. “Specifically, people who could not really afford a human financial advisor because they did not have the assets to pay them.”
Software companies began to move in two directions. Developers either looked to produce fully automated systems, or designed hybrid programs where consumers would receive initial advice through AI — and, later on, could seek out more specific help from a human advisor.
“Often, what people really need is advice to avoid making the biggest mistakes, like letting money sit in your savings account while your credit card debt gains interest,” he said. “Once you fix those, you’re basically 90% of the way there.”
Ho echoed Rossi’s views, emphasizing that Intuit is not looking to replace human financial advisors, but rather seeks to use them to bridge that remaining knowledge and advice gap.
“It’s not about reducing human interaction, but instead increasing the quality of those interactions,” she said.
Thus, instead of paying a financial planner from the beginning, users could save money and time by using generative AI for the bulk of the necessary preliminary research and advice a financial planner would likely charge significant fees for.
“AI systems are great, but they’re not 100% right all the time — so that’s where you pair it with a person who has years of that domain expertise so you can take that next step and act on the advice it gives you,” added Ho.
While the jury is out on robo-advising as a way to make financial planning more inclusive, Rossi said, he said it’s important that these programs be designed in a way to ensure individuals are willing and able to divulge the necessary information to render useful advice.
In particular, “soft financial information” — like how volatile one’s employment status is, or the average life expectancy of one’s family — may be easier to obtain through one-on-one interactions with a human advisor a customer feels they can trust, as opposed to an unfamiliar AI program.
“People can be very secretive about their financial information,” said Rossi, “so what people need is a holistic robo-advisor that takes into consideration the entire picture, and is able to get that picture based on the kinds of questions it asks and how it asks those questions.”
Perhaps most important when considering what financial AI to use, added Rossi, is how the software protects and uses one’s data. He cautioned against software developers who may use personal information (like age or gender) to market specific products through the advice it generates, like promoting one bank’s line of credit over another as a paid advertisement under the guise of unbiased advice.
Ho also noted the importance of doing one’s research before selecting which personal finance AI to use. She said consumers should be careful to use generative AI specifically designed for financial planning rather than any old chatbot out there.
“Data privacy is incredibly important, but so is understanding that not every generative AI system out there is created equal,” she said.
“What we don’t want to happen is folks who already don’t have great access to financial resources using systems that give them terrible advice and put their data at risk — so it’s not just about financial literacy, but data literacy, too.”