A version of this article first appeared in CNBC’s Inside Wealth newsletter with Robert Frank, a weekly guide to the high-net-worth investor and consumer. Sign up to receive future editions, straight to your inbox.

Market data firms have been pitching artificial intelligence as the key to locating elusive ultra-high-net-worth clients. But leaders at elite advisory firms told Inside Wealth they aren’t sold.

For starters, while AI products can surface data and contact information on ultra-high-net-worth individuals, that’s only half the battle.

“When we’re looking for clients with north of $100 million, I struggle to think they’re going to take a cold email and say, ‘Yes, here’s my balance sheet,'” said Matthew Fleissig, CEO and co-founder of Pathstone, a registered investor advisory with $182 billion in client assets.

Instead, he said referrals come when the company works on a more personal level, like when Pathstone once secured a private jet in under an hour for a client who needed to get from New Orleans to Albany, New York, before their mother died.

“Those types of things are how we are able to grow the business,” he said. “We create moments that matter.”

Fleissig said AI for client prospecting hasn’t been the gamechanger that startups purport it to be.

“These databases have been around forever, and now people have added an AI overlay to be able to mine the database,” he said. “Most of the time, it’s very similar strategies of aggregating data sources that are public or you can pay for, and trying to feed you lists of people. We, at this point, can do that ourselves.”

A growth executive at a high-end national RIA told Inside Wealth that he had done at least 20 demos of AI client prospecting tools in the past six months and said most are built on widely available large language models like Claude and GPT.

“You’re slapping a coat of paint on one of five major LLMs and selling through the fact that ‘Oh our info is better,'” said the executive, who requested anonymity to talk about client acquisition strategies. “Do I pay them $100,000 or do I talk to my IT team and figure out a way of doing it for cents on the dollar?”

Andrew Douglass, head of growth at AlTi Tiedemann Global, said there is little competitive advantage to using nonexclusive data. When the independent wealth management firm used to cold call clients from these types of databases, the client usually already had an advisor or had been called by dozens of other firms already, he said.

For the past five years, client referrals and personal networks have made up 40% and 30%, respectively, of AlTi’s organic growth, he said. Another 30% comes from networking with experts like trusts and estates lawyers and accountants who are likely to be working with clients going through a liquidity event, such as inheriting a fortune or selling a business.

“Most people go out and say, ‘Our minimums are $25 million so whoever has $25 million in liquid assets makes a great client.’ We don’t think that that is a strategy that ultimately works,” said Douglass, calling from the Heckerling estate planning conference in Orlando, Florida. “We think really being looked at in the market as a subject matter expert, consistently showing up to places like Heckerling and where the professional community is and being able to provide value, is the most effective way to grow the business,”

Word-of-mouth referrals are not inherently scalable and can be slow-going. Douglass said the sales cycle with an ultra-high-net-worth client can take 12 months, if not longer.

However, advisories focused on the ultra-rich like AlTi Global are looking for quality, not quantity, he said. The firm’s annual target for organic growth is 25 to 30 new clients in the U.S., which could add about $1.5 billion to $2 billion in new assets.

Eden Ovadia, CEO of AI client prospecting startup Finny, said she is used to encountering skepticism. Ovadia, who co-founded Finny in late 2023, said she views AI prospecting as a complement to traditional outreach rather than a replacement.

She said a popular way for high-end advisors to use Finny is to promote exclusive events to the right audience. For instance, an advisor looking to invite prospects to a suite at a Miami Heat game can use Finny to identify people who work in real estate and are interested in the team. Ovadia also said Finny can be used to identify clients who might need advice after a life transition, such as finding people who recently bought a property worth at least $5 million near Jackson Hole, Wyoming.

“There’s definitely a little bit of cynicism we have to get over when we talk to ultra-high-net-worth firms and they’re, ‘No, we don’t do AI. We want everything to feel really personalized, really white glove,'” she said. “I couldn’t agree more. The idea here is we actually can surface more data about your clients or your prospects than even you know.”

Finny can also be used to keep an eye on existing clients and monitor for signs they may be unhappy, such as searching for investment advice online, Ovadia said.

Fleissig said he is more excited about customers finding Pathstone through AI platforms like Gemini and ChatGPT. In the past two weeks, he said, Pathstone has received five inbound inquiries from clients worth at least $100 million from AI search engines.

Douglass said while AI hasn’t changed the way AlTi Global finds new business, he’s open-minded.

“If someone has a better mousetrap, we’re certainly excited about what the market’s going to look like and bring to bear,” he said.