Advisors Use AI for Email, and Vanguard Knows Why

The most common use of artificial intelligence in an advisor's day, according to Vanguard's new survey of the profession, which polled 549 U.S. advisors, is drafting an email, cited by 38%, with research at 35% and meeting notes at 27%—the three most-cited applications all sitting in the same modest category of sentence-length output a human can review before it leaves the building.

Adoption, on this evidence, is real and shallow at once, and Lauren Wilkinson, who runs information technology for Vanguard's financial advisor services unit, put the condition plainly in a statement accompanying the research: many advisors have only begun to scratch the surface of AI's potential, and firms will need to move from using the technology to assist with tasks to using it to automate them. The distance between assisting and automating is not a product feature; it is a supervisory decision, a liability question, and a workflow rebuild, and the barrier data in the same survey shows precisely where that distance gets spent.

The leading obstacle was compliance and home office hesitance, cited by 37% of respondents, ahead of the 34% who said they lack time to learn new capabilities and the 31% who cited limited proficiency; the fourth barrier, named by 22%, is the one advisors tend not to say out loud in front of clients—the worry that AI undercuts their own value. Three of those four are training problems addressable with onboarding and templates; the largest is a permission problem. Letting a model produce something that reaches a client unsupervised requires a supervisory file a chief compliance officer will defend under examination, and no product demo solves that. The person who signs the review, rather than the advisor or the principal, is therefore frequently the effective buyer of advisor AI.

Underneath all of it sits time: 72% of advisors surveyed said they wished they had more time to prospect and deepen client relationships, and Vanguard's answer is already inside the same document—model portfolios, which the survey found leading among portfolio management solutions. Eve Cout, who heads advisor solutions for Vanguard's advisor services business, said in the same statement that research suggests advisors can save more than 400 hours a year using model portfolios, and that the value lies in what they do with the time; she is among those attending this week's Future Proof Festival in Huntington Beach, Calif.

The portfolio data complicates the outsourcing case more than the headline does. Investment performance and track record ranked as the single most important consideration for 68% of respondents, cost for 51%, and 32% of advisors still manage portfolios entirely on their own. An advisor who hands over model management while ranking performance first is a tenant on a short lease: the decision gets made on returns, reviewed on returns, and reversed the first year returns disappoint. Outsourcing wins back the hour; it does not win the client. That is the constraint on every model portfolio shelf, Vanguard's included, and it makes the 400-hour arithmetic an argument for trying the product rather than a reason to stay on it.

The survey is doing two jobs at once, and Vanguard does not pretend otherwise: it sells advisor solutions, and a survey that opens on a time shortage, names compliance as the obstacle, and closes on model portfolios is routed toward that shelf. Naming the alignment of interest is less an accusation than a catalogue entry—this is now the standard form of the genre, asset managers publishing adoption data that routes an operational gap into demand for their own shelf.

The client-facing layer of AI is the least adopted of the three applications—meeting notes at 27%—a figure that lands five days after Escalent's Brandscape research, covered in this publication, found advisors spending 59% of the week on relationships and 34% on portfolios with AI adoption at 68%. The same research house, Escalent, fielded Vanguard's July sample through its Cogent Beat Advisor panel, so the two datasets are looking at the same population from different angles and agreeing on the shape of it: hours that exist, hours that are billed, and a technology that so far has been pointed at the gap between them.

As this publication has argued, the AI-wealth fight is a distribution war, and the winners will own the plumbing and the advisor relationship rather than the model—a view Vanguard's figures are consistent with and sharpen. The plumbing is going in at the task layer, where 38% of advisors have already conceded the least valuable minutes of the day, and the relationship layer—meeting prep, follow-up, the conversation itself—remains almost untouched at 27%. The model layer, meanwhile, is priced: performance first at 68%, cost second at 51%, a third of the market still doing the work by hand, and free to keep doing it.

The gap between those two applications is the number to watch. Email drafts and meeting notes are tasks the advisor already owned and merely delegated; once note-taking and pre-meeting prep pass the email draft, the technology has moved into the hour clients pay for, and the compliance question stops being about efficiency and becomes about supervision. Until that happens, a 38% adoption figure for email drafting means the industry has automated its least consequential quarter hour, and the 32% still managing portfolios on their own have not yet been given a reason to move.