Morningstar survey finds 80% of advisors use AI, but client-facing time stays at 53%

Morningstar's newest reading of the profession finds the ideal client-facing share of a workweek at 63% and the actual figure at 53%, a ten-point shortfall the firm translates into four hours out of a 40-hour week. That gap is the setting for the survey's louder number: 80% of U.S. advisors now use artificial intelligence in some form, up from 67% a year earlier.

The 2026 Morningstar Investor Perspectives Advisor study, reported Monday by InvestmentNews, drew 501 online responses between July 21 and Aug. 16 from advisors at independent and regional broker-dealers, RIAs, wirehouses, insurance broker-dealers and other financial institutions. The sample is not RIAs alone, and "in some form" is a wide threshold that the survey's own task list illustrates, running from meeting summaries and email drafts to brainstorming sessions and client messages. Adoption has nonetheless run to near-saturation: non-users fell from a third of respondents a year ago to a fifth.

Sentiment moved with usage, though not to consensus: 42% now describe AI as a help to their practice, up from 33%, while 19% call it a threat and 39% are undecided. The largest single group has not made up its mind, which is a strange place to be for a technology that four-fifths of the respondents already touch.

The evidence that the tools deliver is not thin. Forty-eight percent of advisors say AI has significantly or moderately improved their efficiency, up from 36% the previous year, and 57% name more efficient client communications, in the form of note summaries and follow-up messages, as its biggest single contribution. The puzzle is what that improvement has not bought: if advisors are running faster, the share of the week that reaches clients should be climbing toward the 63% they call ideal.

Joe Agostinelli, Morningstar's senior director of market research, explains it this way: "AI is helping advisors run faster, but the treadmill is speeding up too," he said. "More documentation, faster service expectations, and more client questions mean AI is absorbing complexity rather than simply freeing up time."

The ranking of obstacles supports that account. Fifty-six percent named administrative and operational work as the top barrier to delivering the service they want, well ahead of a general lack of time at 37%. Advisors are mainly saying the hours they have go to the wrong work, and that the AI now in place has absorbed that work rather than removed it.

Forty-three percent use AI for internal tasks such as meeting summaries and email drafts, 36% for brainstorming, 35% for research and due diligence, and 33% for client messaging. Multiple answers were allowed, so those shares overlap and one advisor can appear in three of the four. All of them sit in the preparation and follow-up around a decision rather than on the decision itself. The reliability verdict follows the same pattern: only 18% rate AI tools as highly reliable for investment recommendations or portfolio decisions, a defensible place for a fiduciary to hold the line and a ceiling on how much client-facing time the technology can return.

Morningstar is not alone in this finding: Vanguard research in September concluded that most firms are using the technology for administrative work. Platform vendors have spent the past year pointing their AI directly at the advisor-client conversation, among them Vanguard's tax-AI deal with Altruist, LPL's Latitude build, and the meeting tools from Practifi and AssetMark. PWD has argued the payoff from all of it runs through the client meeting. The survey supplies the demand-side answer: advisors are using what they have been handed, mostly for drafting and summarizing, and rating it far higher for communications than for portfolio work.

The supply side has been building to a different timetable, and LPL's Latitude launch lifted the stock 3.2% the day it landed. None of that has yet shown up in the part of the week advisors say they want back.

The study measures what advisors report, and it leaves an obvious question open: whether the four hours would appear if the operational load were actually cut, or whether the appetite for advisor attention simply expands to fill whatever capacity is freed. Agostinelli's treadmill is that hypothesis in one image, and a client-facing share stuck at 53% is consistent with it.

For an RIA principal, the useful part of the survey is not the adoption figure: at 80%, using AI no longer separates one firm from another, and the fifth that abstains is not where the competitive question sits. What Morningstar has measured is a practice whose bottleneck moved: the tools deleted some drafting and summarizing, and the complexity surrounding the work expanded to fill the space, with more documentation, faster response expectations, and more client questions.

That suggests the next increment of client time will not arrive with another license. The barrier advisors ranked first, the operational and administrative load named by 56%, is the kind that yields to redrawing who does what, how client data moves between systems, and which tasks a firm decides to stop doing. It is unglamorous work that registers in no adoption statistic.

Morningstar's estimate of the prize is four hours a week, roughly a tenth of the working year, and the survey's own ranking of obstacles puts those hours in operational work rather than analysis. Adoption has little room left to climb. The 53% is the number next year's study will have to move.