Schwab bought AI access; Altruist owns the client record

Schwab Advisor Services used Future Proof to put a product name on its AI ambitions, unveiling an exclusive arrangement with Anthropic that would bring Claude for Financial Advisors to the 16,000 RIAs that custody and run back-office operations on its platform, a debut RIABiz reported as an answer to Altruist, the AI-native custodian Vanguard is set to own. The two analysts the outlet consulted returned a number every RIA principal weighing a custody decision should sit with: two to five years, they said, is how far ahead Altruist remains on the agentic tooling advisors are being promised.

The sharper evidence came not from the analysts but from Jason Wenk, Altruist's founder and CEO, who pointed to Arca, an RIA with more than $1 billion in assets, leaving Schwab and consolidating on Altruist exclusively — a switch he attributes to owners who can see where the technology is heading rather than to price or service. PWD's records log Arca's custodian change on September 10, and the move matters out of proportion to its size: a billion-dollar book is the kind of account an incumbent works to keep, and its owners decided the workflow around the account had become the reason to move it.

Lex Sokolin, managing partner of the London venture firm Generative Ventures, describes the gap as compounding, with Altruist's advantages, in his telling, cumulative and snowballing as the technology matures, a dynamic he says reprices everything and can produce winner-take-all outcomes. Alois Pirker, founder and CEO of Pirker Partners in Marblehead, Mass., reaches the same place from the consulting side, writing in an email exchange that AI momentum across the RIA business, Altruist's positioning as a thought leader on the subject, and its landmark deal with Vanguard amount to a perfect storm for the established custodians.

Neither man is claiming the incumbent loses. What they are describing is a contest fought on a unit most custody executives have never had to compete on. Custody has been a business of switching costs: the deeper an advisor's book is woven into a platform, the more expensive the exit, and the less a rival's feature list settles anything. AI-native software attacks that wall at its thinnest point, and in an AI-native stack, the integration is the product.

The bear case for Altruist is arithmetic, and RIABiz states it plainly: the startup holds roughly as much as Schwab brings in during a good week, and against a $6-trillion custody book, that is not a disruption; it is an experiment, with the four custodians RIABiz names alongside Schwab — Fidelity, Pershing, State Street and Apex — staying in the game for the incumbent's reason. Once the plumbing exists, a marginal advisor costs almost nothing to serve, and defending share is a fraction of the price of winning it.

Our desk's bias is that scale is oversold and durability underpriced, and this fight is the cleanest test of it available. Schwab's real asset is distribution: 16,000 advisors, a $6-trillion book, and a brand that took decades to assemble, none of which a software release erases. Altruist's real asset is that nothing it ships is a retrofit, and when the technology underneath an industry turns, the incumbent's edge is reach and the challenger's edge is the shape of what it already built, reach counting for less in a market where an advisor can sit through a competitor's demo on a Tuesday afternoon.

There is a reason custody changes are rare, and it is not satisfaction. Repapering accounts, retraining staff on a new trading and reporting stack, and rebuilding the integrations a firm depends on all land in the same quarter, and the disruption is absorbed by the advisor's own clients. A platform that lowers the operational cost of the move changes that math directly, which is the mechanism by which a technology lead becomes a market-share number.

The timeline deserves pressure. RIABiz's own note observes that in AI three months is an eternity, a line that cuts both ways: Altruist's lead is wide on the clock of model releases and narrow on the clock of custody contracts, which turn over on the timetable of a principal's patience. What the estimate actually measures is how long it takes to replicate a workflow that was designed around AI from the first line of code, and nobody — including the analysts who offered a range — can put a confident number on that.

The piece of the contest the product names obscure is data. RIABiz reports that RIAs are caught in the middle over data concerns as the soon-to-be Vanguard-owned custodian and the $6-trillion incumbent trade shots, which is the right place to look because, as this publication has argued, the AI premium has moved from model quality to the governed client record: whoever holds the permissioned copy of the client data is the party the meeting runs through, and model quality is the one input in this stack that reliably converges.

Read that way, Schwab's Anthropic deal is best understood as a queue-position purchase, an arrangement that gives the custodian an official claim on Anthropic's attention and a Claude for Financial Advisors label to carry into every conference from here. What it does not hand Schwab is the connector layer that turns a model into a workflow — the part of the stack where Altruist built first, and the part a licensing agreement lets a firm reach without owning.

Anthropic gets something concrete from its side. A custodian with 16,000 RIAs is a distribution channel into advisor workflow that no direct sales force matches, and the exclusive label tells the market which model the incumbent has chosen. Which party that arrangement is worth more to will price the next round of these partnerships.

There is also a vendor-risk question that neither announcement answers. A custodian that threads a third-party model through its advisor workflow inherits that vendor's roadmap, pricing and availability as if they were its own, and any repricing of model access lands on the platform's economics before it reaches the advisor. That is the standing argument for owning the connector rather than renting the last mile, and it applies to the incumbent as squarely as it applies to a startup.

The data anxiety RIABiz describes has a practical shape. An advisor's client record lives inside the custodian's systems, and the more the platform does — billing, reporting, planning, and now AI — the more of a firm's operating life is stored somewhere the firm does not control. RIAs that spent the last decade consolidating their own systems are now being asked to stack their intelligence on top of someone else's.

Ownership complicates every projection: Vanguard is set to take control of Altruist, and it has already hired a former Barron's editor, Beverly Goodman, to bridge its low-fee message to the advisor-sold channel. A low-cost asset manager that owns an AI-native custodian and is building an advisor distribution arm is a harder competitor for Schwab and Fidelity than a venture-backed startup with a head start, because it puts fee compression and workflow software in the same house, aimed at the same advisors.

RIABiz reports that RIAs with $1 billion or more in assets and an appetite for technology are starting to join the platform, which suggests the lead — whatever its exact length — is converting at the top of the market rather than the tail, the segment the large custodians can least afford to lose. Those accounts carry the referrals and the economics that fund everything beneath them, and they are the accounts most likely to run a formal custody search, which is exactly when a demonstration stops being a demonstration.

For an RIA principal, the practical question is who holds the client record, what it costs to extract it, and how much of the workflow the custodian actually controls when the model underneath changes, not which platform has the better demo. Those terms are negotiable at signing and rarely negotiable afterward, and they deserve more weight in a custody RFP than any keynote. A principal who cannot answer them has handed the decision to a roadmap.

The call is this: Schwab bought visibility, Altruist is spending on durability, and the two-to-five-year estimate flatters the giant only if Schwab converts a model partnership into owned workflow faster than Altruist converts a workflow lead into contracts at scale. Nothing in the reporting so far establishes that it can. The next handful of billion-dollar moves, in either direction, will settle whether the head start was worth five years or about one good week.