Family offices are increasingly investing in artificial intelligence but have been slow to adopt it within their own operations.
Over half (53%) of family offices have built up portfolio exposure to generative AI technologies, with an additional 26% considering it, according to Citi Private Bank’s 2024 Global Family Office survey.
AI is becoming a central focus for private wealth management firms, with former UBS CEO Ralph Hamers recently joining the AI-driven startup Arta Finance.
But only 15% of family offices surveyed by Citi were using AI for operational purposes — such as automating tasks, building presentations and forecasting. “Few inroads have been made when it comes to potentially more value-added applications such as portfolio construction and risk management,” the Citi survey said.
That dichotomy can be attributed to the challenge of overcoming long-standing family office habits and practices, taking time to choose the best uses of a constantly evolving technology, and concerns over compliance and cybersecurity, say family office members and AI providers.
‘Much more efficient’
The uses of AI in family offices range from front-office tasks such as portfolio management and analytics to back-office tasks such as accounting and communication.
“AI has made tasks much more scalable, much more efficient,” said Danny Lohrfink, the co-founder and chief product officer of Wealth, a digital estate planning company based in Tempe, Arizona.
“And at the end of the day, that’s what technologies tend to do — and when you drive scale and efficiency, you tend not to sacrifice on quality. But you do end up saving time and money.”
Lohrfink cites the example of one family office in Florida that “effectively embedded AI on the data layer and all the underlying data points” to track all the places that a family member was mentioned in its complex estate plan. “That is something that would have taken an analyst within that office days or weeks to comb through all those documents and data from a variety of programs and platforms,” he said.
Benny Clavero, vice president of client services at Keebeck Wealth Management in Chicago, said his multifamily office uses AI to automate personalized messaging and portfolio reviews for clients.
“Rather than going through each one, reviewing it and giving a summary to the clients, we’re going to have the portfolio review tool read it and then send them [the summaries] to individual clients automatically,” Clavero said.
That move has helped streamline operations and reduce costs, he said. “We’re basically eliminating operations, removing the need for a client service associate, so it basically goes up to the manager or supervisor level.”
Clavero’s firm is also starting to use AI for investment purposes “to analyze our entire portfolio and notice any trends and then review the entire book” to look for deviations from where the investment was initially allocated and then make recommendations.
“In the end, obviously, there is a human who makes the actual decisions and pulls the trigger,” he said. “The AI basically just does all the data entry.”
Investment assessment
AI’s sheer processing power can also be used to screen potential investments and for benchmarking, said Jake Miller, co-founder of the private markets investing platform Opto Investments.
“A manager will send you a 30-page deck and a nine-page private-placement memorandum and other supporting documentation — how do you distinguish yourself from the next investor without having to pore through every page?” he said. And even if you don’t end up investing, all those materials “feed into your system to grow your view of the universe,” Miller said.
It’s also helpful when assessing performance. For example, Miller said, “this manager is better than 80% of the managers we saw in the last six months on fees and better than 50% of them on performance.” That quick assessment by AI can help family offices save hundreds of hours on screening managers so they can focus on doing deals.
The use of AI was transformative for one of Miller’s clients, which had been experiencing “real pain points” by overestimating total calls and underallocating people. The firm started using AI last year and in its yearly reviews last month and found it to be “a completely different experience,” he said.
Security concerns
But with the benefits of AI also come increasing cybersecurity threats, said Leo Schrantz, chief AI and innovation officer at Wakefield, Massachusetts-based Vestmark, a provider of wealth management software.
“You need to be very careful about where your data’s going,” Schrantz said. When family offices deal with AI vendors, as part of their due diligence, they need to have the vendor walk through the chain of custody, he said: “Who’s your large-language-model provider? Where is that hosted? What’s your relationship with that provider? Do you have the right privacy and protection?”
Most investors are underprepared for the wave of generative AI scams or frauds that are coming in the next few years, Schrantz said, “specifically with the rise of avatars and replications. You get a phone call and might assume you’re talking to your kid who needs bail money or your financial adviser.”
Regulation speculation
Another hurdle to the widespread adoption of AI is the uncertainty of the compliance landscape.
Broker-dealer LPL Financial uses AI to help advisers provide personalized insights to clients, but it is “moving cautiously, since regulatory authorities like the SEC are still looking into the use of AI in portfolio management,” said Kristie Edling-Day, the firm’s chief information officer and executive vice president.
Keebeck Wealth’s Clavero said his firm limits the capabilities of its AI when interacting with clients. For its chatbot, “we have set parameters as far as where it sources that information as well as the kind of responses that it can give” so that it’s not taking on the role of a financial adviser and giving buy and sell recommendations, he said.
The promise of the tools can also lead to overconfidence and an eagerness to adopt technologies “you don’t totally understand because you don’t want to be left behind — classic FOMO [fear of missing out],” Miller said.
Successful use of AI depends on family offices realizing that it “requires data capture and thinking about how you want to set it up,” Miller said, especially if you’re still using paper and handshake agreements.
“You have to be willing to change the way you do business to build the data sets to have the AI transform how you work through problems.”