RIAs are hiring around AI before they can manage it

Cerulli Associates asked 68 RIA firms what they intend to hire over the next two years, and the answers sorted themselves by proximity to a client: junior advisors topped the list at 73% of respondents, client associates followed at 67% and senior advisors at 56%, while administrators drew 23%, marketers 18% and compliance specialists 15%, according to the survey fielded in May and June and released this week.

That spread matches what the industry has predicted generative AI would do to advisory shops—firms are adding the people who sit across from clients and holding off on the people who process the paperwork. Asher Cheses, Cerulli's senior director of wealth management, told Financial Planning that administrative and back-office roles are effectively on pause at many firms while they work out which tasks the technology can absorb. The hiring decision waits on the answer.

By itself, the list describes a business confidently reallocating headcount toward revenue, but against the same survey's second half it describes a bet placed ahead of the evidence; Cerulli built an AI Maturity Score with Vista Equity Partners, the technology investment firm, scoring each responding firm from 0 to 100 on governance and readiness, operational efficiency, and innovation and revenue.

The average respondent scored 32. Half the firms landed in the "exploring" stage, meaning they use AI but not in a managed way, and many lack written policies on acceptable uses, training in the technology, or data formatted for AI, with employees experimenting individually rather than firm-wide. Thirty-eight percent were "scaling," typically applying AI to client communications, meeting preparation and note-taking with use policies established or underway, while 12% were "leading," using it across advisory and back-office work.

For that 12%, holding off on an administrative hire is a strategy because the tasks have moved and the seat does not need filling, while for the half still exploring the same pause is a placeholder, since the work has not moved anywhere and nobody has been hired to do it. The distance between those two positions is what the maturity score measures and what the hiring plan quietly assumes.

Cheses said the team went in expecting the largest firms, the ones with the biggest technology budgets and staffs, to be the most eager adopters; what the survey found about size and scale was not what they had expected. Take budget out as the driving variable and maturity becomes an operating problem—policy, data and supervision—which suggests a small practice with its house in order can outrun a much larger one, and no firm can write a check that stands in for the work.

Fifteen percent of surveyed firms plan to add compliance specialists over the next two years, at a moment when half of all respondents are using AI without managing it and many have no written policy governing what the technology may do. The function that decides which uses are permitted, what gets reviewed before an output reaches a client, and how the record is kept is the one firms are least inclined to staff; the survey's two halves contradict each other here, with the population betting its back office on automation not planning to hire the people who would make that automation defensible.

The durable AI edge is not adoption speed but data custody, audit trails and proof built over years, and the firms that own the review step keep the clients. The review step is a person with a policy, a log and the standing to say no. Fifteen percent is a thin plan for the function that determines how much of the promised automation a firm is actually allowed to use.

For a firm in the exploring cohort, the sequence is unglamorous: name the administrative tasks in writing before pausing the search for whoever did them, because a pause without an inventory is just a vacancy; get the acceptable-use policy on paper before the client-facing hires arrive, since a 73% appetite for junior advisors means more people generating notes, drafts and client correspondence that someone has to review; and put the data in a shape a model can use, which is the piece the exploring half is furthest from.

None of that argues against the hiring itself, since the client-facing roles on the list are the ones AI has not reached and adding them adds capacity to the work that pays. Fidelity's 2026 outlook makes the point from the other end, arguing that AI's real payoff is time for the planning conversations wealthy clients value most, and the widening gap between heavy and light AI users is already visible in operating results elsewhere in the field.

There is a second gap in the same table: 73% of firms plan to add junior advisors and 56% plan to add senior ones, and junior advisors need senior advisors to train them, which takes the very hours the automation is supposed to hand back. If the average firm is still scoring 32 when those junior hires arrive, the training load falls on the partners who were counting on the technology to free them, and the compliance line stays in the teens while the client-facing roster grows.