We are in the middle of month two of what many of my fellow technology prognosticators and I are calling the year of agentic AI, and look at the selloff in legacy financial services stocks we experienced this week.
Many financial analysts attributed this to the still relatively young custodian, Altruist, and this week’s rollout of its first AI agent, dedicated to tax planning, which is but one part of its Hazel AI platform.
Altruist CEO Jason Wenk was asked on CNBC on Thursday whether he was surprised to see the market reaction.
“We were certainly excited about the release of Hazel, we understand the power it has for financial advisors and their clients, but I’d be lying if I said if I thought I’d check the markets and see $20-plus-billion dollars market cap wiped out in a day,” he said, noting that this figure appeared to have grown in day two. “So, definitely a faster reaction than anyone could have predicted.”
Keep in mind that Hazel just came into existence as such about six months ago, born out of Altruist’s acquisition of AI-powered productivity platform Thyme in June (with a great deal of subsequent additional development, of course).
I’m not a stock analyst, but I have covered advisor technology for almost 20 years, and the example that popped into my head to illustrate the comparatively slow pace of change when we talk about legacy custodians is account opening.
Let’s just take that single process at one custodian, Schwab. The first time I wrote about innovation in account opening was in 2010, with the announced pending integration of Laser App’s capabilities, which would enable it to pull CRM data into and fill out account-opening forms, and speed up the process via Schwab Advisor Center.
The firm continued to work on automated accounting opening with partner Orion in 2018 and discussed killing forms and moving to a fully digital account opening process in 2019, rolling out what it called a fully digital account opening in 2021, with significant updates in 2022. And Schwab was certainly not alone; Fidelity had announced its initial digital rollout in 2020, as did wirehouse Merril Lynch in 2021.
While Altruist launched and began life as an introducing broker/dealer in 2019 and relied on third parties for much of its backend processing for its first few years, it achieved fully digital, same-day account opening and funding when it became self-clearing in 2023.
To be sure, digital account opening is a very complex problem both in terms of technology and the many security- and compliance-related regulatory requirements.
But what brought the digital account-opening process example to mind in the first place was an announcement this week by the startup Dispatch. Founded in 2022 by three co-founders with a mix of technology and financial services industry experience, they have been building what they call a universal data orchestration layer for wealth management.
Schwab had recently created an API-based account onboarding, and Dispatch is one of the first third-party providers to take advantage of it. That means, in the simplest terms, that Dispatch users can kick off an account-opening workflow by enabling firms to send data directly from their CRM or other systems into Schwab Advisor Center, shortening account-opening times and eliminating the need to leave Dispatch to open an account.
For my purposes, it is an example of the kind of continued evolutionary innovation and workaround we will see, keeping legacy custodians relevant for years to come.
I surveyed several other industry technology startup leaders I had spoken with recently about the significance of agentic AI in relation to Altruist’s news and the market selloff.
“In my opinion, what the market is actually reacting to is the growing belief that AI will begin to reshape core wealth management workflows—planning, tax optimization, reporting and compliance—most of which are still driven by manual, billable processes today and quite frankly terrible user experiences for the advisor, and this is one indicator of that disruption coming to the market,” said Sam Sova, co-founder and CEO of Subatomic.
Subatomic connects a wealth management firm’s existing tools so they can talk to each other, then builds AI co-workers on top, teammates that are trained on the firm’s unique processes and internal knowledge and that are meant to be ready to work across departments from day one.
Sova, who spent several years in the financial services industry at TIAA and Fiserv, said that the more important signal is strategic rather than tactical, illustrating how AI is moving from being an assistant to becoming an operator inside financial workflows as it gains the ability to reason across household data, execute compliant actions and learn firm-specific processes, the economics of wealth technology will shift toward whoever controls intelligence within the advisor workflow.
“That’s the real story the market is sensing—even if Hazel itself is simply the catalyst of this disruption with one major use case,” he said.
Anna Joo Fee, an attorney and the founder and CEO of Goodfin, an AI-driven, direct-to-accredited-investor private markets platform focusing on late-stage, pre-IPO and venture/growth private equity opportunities, has also been tackling longstanding industry inefficiencies.
“The market reaction is about uncertainty—the long-term shift toward AI-augmented financial services is only just beginning, and I think at first we will continue to see some skepticism from inside the industry while outsiders who analyze markets and see longer-term value creation will start pricing in the risk and costs of older business models,” she said.
Another person I sought some insight from, especially on the tax tool front, was Kevin Knull, a CFP and the CEO of TaxStatus, which recently added two productivity tools to its platform that are designed to help financial advisors improve on accuracy and efficiency of tax preparation ahead of the 2025 filing season (as a reminder, he was also a former CEO of MoneyGuidePro for several years and one of Wealth Management’s Ten to Watch in 2026).
“It’s complicated, and it’s changing all the time,” he said of the tax code, regulations and oversight.
“AI outputs are only as strong as the underlying data,” said Knull, noting that in his own experience, a major challenge for those preparing tax returns are that clients often forget to provide them data, which has an impact on the return and requires the preparer to spend a lot of time going back to get that data, which is why so many extensions are filed.
And then when talking about AI, there is the need to make sure there are guardrails up both to help prevent hallucinations and to provide for the transparency advisors, and CPAs will expect so that they can easily follow the logic behind decisions or strategies, and in turn explain it to clients or if they themselves are audited.
“It is all about the knowledge first, do you have the right data, and do you have those decisions modeled and guardrails, and these are conversations I’m not seeing enough people having right now,” Knull said.
