OpenAI Took Wall Street; Anthropic Just Took Wealth — Schwab's $5.5T Custody Deal Hands Claude the RIA Channel

Charles Schwab and Anthropic announced on September 14, 2026 that Schwab Advisor Services will be the first — and currently only — RIA custodian integrated with Claude for Financial Advisors. Independent RIAs that custody at Schwab can authenticate from Claude into the Schwab Advisor Center, then run connector-driven workflows across CRM, custody, portfolio reporting, financial planning, estate planning and meetings. The package includes MCP connectors, a library of advisor skills, and pre- and post-meeting agents with admin-facing audit logs.

The release frames Claude as a layer that sits over the existing RIA stack rather than replacing it. Use cases called out — meeting prep, financial-plan updates, portfolio-change explanations, drafting follow-ups and updating CRMs — are the exact tasks Schwab's own 2026 RIA & AI study says advisors want AI for, but 63% of firms are still in the experimentation phase. Schwab is effectively seeding Claude into16,000 firms at the moment those firms are shopping for a default AI workflow.

On September 10, 2026, OpenAI introduced ChatGPT for Financial Services, a finance-tuned enterprise product powered by GPT-6 Astra and built with Morgan Stanley and Evercore as design partners. Native data feeds come from LSEG, Daloopa and PitchBook; the tool is built to take over equity research, modeling and pitchbook work — i.e. the labor of junior investment bankers and research associates. OpenAI called the launch out as 'targeting some of Wall Street's most labor-intensive tasks.'

Schwab and Anthropic are solving opposite problems with the same deal. Schwab's Advisor Services franchise has grown AUM faster than its revenue: assets receiving ongoing advisory services hit $6.67T at June 30, 2026, up 23.0% year over year from $5.43T, but Advisor Services asset-management and administration fees of $476M for Q2 2026 grew only 11.7% YoY. AI-driven productivity is the lever to convert AUM growth into fee growth.

For Anthropic, the deal installs Claude as the default advisor-AI layer in front of the largest US RIA custodian in a single press cycle. Anthropic's annualized revenue run-rate crossed $30B in April 2026 from $14B in February, with roughly 80% coming from enterprise customers. With enterprise accounts already at the top of the funnel, Schwab gives Anthropic its first vertical with built-in custodian-grade governance — exactly the procurement pattern the Anthropic finance-agents playbook has been building toward since the May 2026 launch of the 10-agent library and the connector deals with Moody's, IBISWorld and Fiscal AI.

Wealth-tech vendors long argued their dashboards, planning software and rebalancing engines would be hard to dislodge. The Betterment for Advisors read on the Anthropic plugin ecosystem is more honest: split the stack into interface, intelligence and execution, and the first two are AI-native utilities within12–24 months. Execution — custody, trade routing, broker-dealer standing, fiduciary supervision — remains regulator-bound and operationally hard.

That framing explains why Schwab is comfortable distributing Claude. By owning the authenticated connection to the Advisor Center — the custody system, the trade blotter, the audit trail — Schwab controls the gating function. RIAs that adopt Claude for Financial Advisors are still locked to Schwab custody, Schwab reporting and Schwab supervision. Anthropic monetizes the intelligence layer; Schwab monetizes the execution layer; nobody gives away the interface for free.

Morgan Stanley is the most exposed incumbent despite its lead. AI @ MS Debrief hit 98% advisor-team adoption by mid-2024 and lifted document access from 20% to 80% across a 100,000-document research library — the proof point every other wealth firm cites. Morgan Stanley is also a design partner on OpenAI's new banker product. But it serves employee advisors in a wirehouse channel, not independent RIAs, so the Anthropic-Schwab deal does not reach it. It will need to extend OpenAI-powered tooling deeper into plan drafting, CRM and client follow-up to stay differentiated against a16,000-RIA Claude channel that already has the same model underneath.

LPL Financial, the largest US independent broker-dealer with30,000+ advisors, expanded its Anthropic partnership in February 2026 and already has Claude-native workflows in market. The Schwab-Anthropic deal validates LPL's earlier move and increases switching costs for advisors weighing RIA independence versus an LPL affiliation.

BlackRock is the platform play most at risk on the interface side. Its Aladdin Wealth Auto Commentary uses generative AI to turn portfolio data into client-ready narratives — exactly the meeting-prep and plan-update workflow that Claude for Financial Advisors now ships out of the box for any Schwab RIA. BlackRock still owns the risk-and-portfolio-engineering layer underneath, but the human-facing wrapper is being absorbed into Anthropic's skill library.

Salesforce Agentforce for Financial Services — named #1 financial-services software in G2's 2026 rankings — pitches the same agentic-AI stack to wealth firms but without an integrated custodian. With Schwab backing Anthropic and LPL already on Claude, Salesforce loses the natural distribution wedge into the two largest wealth channels and must either match the connector depth or pivot to advisor firms with their own custody arrangements.

Short-term (days–quarters): the Schwab-Anthropic product will be demoed at Schwab's IMPACT 2026 conference, where independent RIAs make annual custody-stack decisions. Adoption milestones to watch: Morgan Stanley Q3 2026 earnings in October for any update to the98% Debrief figure and any extension into agent-based workflows; Schwab Q4 2026 results in January for AI-driven Advisor Services fee growth versus the 11.7% Q2 2026 baseline; and any move by Fidelity Wealthscape, BNY Pershing or LPL to sign a competing Anthropic or OpenAI deal — Fidelity cleared its own AI product (Wealthscape Intelligence) in September 2025 and would be the obvious next custodian to follow.

Long-term (1–3 years): if Fidelity research is right that AI can lift advisor productivity by 25–40%, the economics of an RIA practice rebase. Today the median Schwab-custodied RIA runs a 30%+ EBITDA margin because advisor compensation is the dominant cost. A 30% productivity gain either lets the firm absorb more clients at the same headcount (revenue up, margin up) or cuts advisor count (revenue flat, margin up sharply). In either case, custody revenue per RIA firm rises — a tailwind for Schwab that compounds with the existing23% AUM growth.

The structural risk is concentration. With54% of US RIA custody already at Schwab, the Anthropic deal makes the RIA channel more, not less, dependent on Schwab's choices. If OpenAI strikes a counter-deal with a second-tier custodian, RIAs may get genuine model choice; if it doesn't, Anthropic's lead at the wealth advisor layer compounds the way OpenAI's lead at the banker layer is compounding now.