According to Vlad Golyk, coauthor of a recent McKinsey report on AI and wealth management, more than a third of consumers across all age groups are turning to tools like Claude and ChatGPT for guidance on their investments. In many cases, he says, they are consulting the tools ahead of meeting their real-life financial advisor.

This trend is only going to accelerate as Anthropic and others find new ways to infuse financial services with artificial intelligence. The full implications are not clear, and it’s easy to see things going wrong. (Picture a future where a YOLO AI advises its clients to go all in on penny stocks.) The early signs, though, paint a more positive picture—one in which financial insight becomes more democratized and AI tools augment, rather than replace, money managers.

What’s old is new—and different

This isn’t the first time we’ve seen buzz around AI and personal finance. Back in 2016, startups began touting AI-powered tools, known as robo-advisors, that could adjust investment portfolios based on age and market conditions.

But robo-advisors are today regarded as a generic, incremental feature at best. The latest versions of AI financial advisors, trained on powerful LLMs, are different. That’s the view of Robinhood, which counts 250,000 customers paying on average $250 annually to use its Strategies, an AI tool guided by human advisors. Product manager Sam Nordstrom says earlier AI tools simply slotted customers into one of 20 or so baskets of ETFs based on a questionnaire, but that the company’s new offering can account for far more scenarios. For instance, it can provide guidance to someone who has the bulk of their wealth tied to stock options, or is looking to build a portfolio that accounts for their plans to buy a home. “Where our team is excited to keep exploring is the technology’s ability to create truly personalized outcomes,” notes Nordstrom.

The good news for Robinhood and its competitors is that they are likely to find product-market fit at a time when human-based advice is becoming harder to find. According to McKinsey’s Golyk, there is a looming advisor shortage as fewer people enter the profession, and those who remain focus increasingly on the wealthiest investors alone. The upshot, he says, is that there is a big opportunity for firms like Robinhood to target with AI offerings those making $100,000 to $1 million.

Human advisors who stay in or join the business, meanwhile, will increasingly be giving advice based on AI-driven models. In February, Anthropic announced a partnership with LPL Financial, which provides wealth management tools to tens of thousands of registered advisors and financial institutions, who in turn advise around 8 million customers. The tie-up comes after the AI giant signed a spate of other deals that allow it to gobble up data from the likes of S&P Global, Morningstar, and PitchBook. All of this means Anthropic is creating something akin to investing superpowers whose wisdom will be parceled out by humans.

Building guardrails

“Jeff, you really do understand the market! If you believe you’ll double your money on that crypto coin, it will happen. Trust your gut and go for it!”

This is an invented quote, but for anyone who has interacted with an AI chatbot, the affirming, obsequious tone will be familiar. Such behavior by chatbots has driven some people to reckless actions and even self-harm, so it’s not hard to imagine the tech egging on an investor to squander their savings.

These sorts of scenarios are why New York State Sen. Kristen Gonzalez is pushing legislation known as “the Chatbot Liability Bill” that would create the right to sue AI providers that impersonate licensed professionals. Discussion about the bill has mostly revolved around doctors and lawyers, but Gonzalez told Fortune it would also apply to unlicensed financial advice.

Gonzalez stressed that the right to sue would apply only to “substantive responses” in which a bot held itself out to be a financial professional. Still, the prospect of such legislation is leading those building personal finance AI agents to err on the side of caution. It is probably no coincidence that Robinhood’s Nordstrom described the personality of its AI offerings as “very objective and factual,” though he didn’t rule out the company reevaluating that in the future.

These concerns are less pressing for banks and traditional investment houses, which not only are more likely to have human intermediaries, but also are working with AI providers to bake robust compliance measures into their AI tools. Morgan Stanley’s head of wealth management, Jed Finn, recently touted how its analysts use AI to “ingest” data about a client before providing advice on topics like Roth IRA conversions or gift-based tax strategies. The company is working with a variety of AI agents to develop market insights that can in turn be paired with AI-curated data for individual clients.

Meanwhile, just as big firms a decade ago responded to the arrival of startup-backed robo-advisors by creating versions of their own, they are now building souped-up bots that offer more natural conversational guidance. That includes Fidelity’s “Freya,” a customer-facing tool that answers personal finance questions but makes very clear its responses are not advice.

The full potential of AI-powered financial advice will not be distributed evenly to all investors. But the early days of the AI financial era are already delivering unprecedented access to a new array of useful financial tools.

For more on AI best practices, visit fortune.com/topic/aiq/

This article appears in the April/May 2026 issue of Fortune with the headline “AI may finally be ready to manage your money. Should you let it?”