We asked five AI systems which RIA to recommend. They produced five different shortlists.
Across six HNW buyer-intent prompts (family-office management, multi-generational wealth, exiting founders, institutional advisory, private-equity coordination, ultra-HNW tax and succession) run three times against ChatGPT, Claude, Gemini, Perplexity, and Grok in May 2026, the five engines named 86 distinct RIAs with an inter-engine Jaccard overlap of 0.32 — lower than commercial insurance brokers at 0.38, and not a single prompt produced a unanimous #1 firm. ChatGPT is the systematic outlier: it tilts toward trust companies and private banks (Bessemer Trust, Rockefeller Capital, J.P. Morgan Private Bank), while the other four engines lean toward independent RIAs (Creative Planning, Mariner Wealth Advisors, Cresset Asset Management). The corroboration apparatus splits between Barron’s and FA Magazine rankings (which ChatGPT and Claude weight heavily) and family-office trade press (which Perplexity and Grok weight more). Compliance note: this measures observed AI surfacing, not adviser quality or fitness.
What this means for you
- 0.32 inter-engine overlap — the lowest cross-engine agreement we have measured. RIA recommendations fragment harder than insurance brokers or PI law (slightly).
- ChatGPT is the systematic outlier engine. It tilts toward trust companies and private banks (Bessemer Trust, Rockefeller Capital, J.P. Morgan Private Bank). The other four engines lean toward independent RIAs (Creative Planning, Mariner Wealth Advisors, Cresset Asset Management). Two distinct shortlist universes for the same HNW buyer question.
- Zero unanimous winners across six prompts. Family-office at $50M+, multi-generational wealth, exiting founders, institutional advisory, PE coordination, ultra-HNW tax/succession — every single prompt produced a different shortlist by engine.
- Implication for your firm: which AI engine your prospective client uses materially shapes which firms enter their consideration set. The Barron's-recognized RIA dominating Claude shortlists may be entirely absent from ChatGPT shortlists, and vice versa. The Diagnostic identifies which engine is the outlier on your firm and which corroboration sources would close the gap. Compliance note: observed AI surfacing, not adviser quality, not investment advice, not solicitation.
Cost of inaction: an HNW prospect researching advisors through ChatGPT is being shown a fundamentally different shortlist than the same prospect using Perplexity. If your firm is absent from one universe, every prospect who starts in that engine starts without you. The lift to be present in both universes is bounded; the lift to recover from a year of one-universe absence is harder. Compliance note: observed AI surfacing, not adviser quality, not investment advice, not solicitation.
For what to do about it, see the Wealth Management & RIAs industry brief →
What this page measures
Each row is not a ranking. It is observed surfacing — how often a company entered the AI-mediated consideration set across a bounded battery of buyer questions. The heatmap maps citation territory: for each question the engines repeatedly surface a small set of companies, and those companies currently hold the answer layer for that question. The question is not "who is best?" — it is "who appears when the buyer asks?"
Observed surfacing, not endorsement. These pages sit inside the same Content / Retrieval / Trust architecture as the rest of our working papers on AI-mediated buyer discovery.
Category temperature: frozen vs molten
Frozen
The engines have converged on a small canonical set, reinforced by broad corroboration, accumulated entity clarity, and a large shared retrieval surface. On-page changes alone are unlikely to displace the top tier; the strategic question shifts to defending edge cases and emerging buyer questions.
Molten
The answer has not set. Engines disagree, adjacent or foreign firms may enter the answer set, and retrieval-surface quality matters more. Companies can still claim citation territory by publishing answer-shaped, entity-clear, machine-retrievable content.
Wealth Management & RIAs is the most molten category we have measured. Mean cross-engine overlap 0.32; zero of six prompts produced a single unanimous #1 firm across all five engines. B2B SaaS & Industrial Manufacturing is the frozen comparison (0.85).
How to read this
Hover or tap any engine cell above to see the real prompts behind that number.
- AChatGPT and Claude lean trust-bank. On the multi-generational-wealth and family-office prompts, ChatGPT and Claude both surface Bessemer Trust, Rockefeller Capital Management, and Aspiriant prominently. Trust companies and private banks dominate these engines — firms with long-tenured wealth-management reputations and corroborating coverage in Barron's, Forbes, and family-office trade press.
- BPerplexity and Grok lean independent-RIA. Same buyer prompts, completely different shortlists. Creative Planning, Mariner Wealth Advisors, Cresset Asset Management, Mercer Advisors, and Pathstone surface on Perplexity and Grok at rates the trust-bank firms surface on ChatGPT and Claude. ChatGPT in particular surfaces zero of the top-14 consensus independent-RIA firms on several of the prompts — the firms exist in our capture, but ChatGPT routes elsewhere.
- Which SEC-registered investment advisors specializing in family-office management are commonly recommended for HNW individuals with $50M+ in assets?
- What independent RIA firms are commonly suggested for managing multi-generational wealth and handling complex trust structures?
- Which regional wealth advisors are most frequently recommended for venture-backed entrepreneurs exiting their companies?
- Which RIA firms are commonly cited for institutional advisory and pension-fund management?
- Which independent registered investment advisors are commonly recommended for coordinating private equity and alternative-investment portfolios?
- Which wealth-management advisors are frequently cited for tax-optimization and succession planning for ultra-high-net-worth families?
Scope — US market, SEC-registered investment advisers ($1B–$50B AUM tier). The six prompts span six HNW buyer scenarios common in family-office and HNW-individual advisor discovery. This is not a comprehensive scoring of the RIA market — large bank-owned wealth platforms and the largest national RIAs were surfaced where the engines named them, but the cohort tilts toward independent RIAs and trust companies in the $1B–$50B AUM range.
Strategic reading: two shortlist universes, no single canonical answer
RIA recommendations fragment because the underlying public-evidence sources fragment. Barron’s Top 100 names one set of firms. FA Magazine RIA Ranking surfaces a different set. SEC Form ADV is the underlying disclosure document but it’s dry and structured, not narrative. Family-office trade publications and HNW-specific media each contribute their own corroboration sets. Each AI engine weights these differently, producing two distinct shortlist universes for the same HNW buyer question: a trust-company/private-bank lean (ChatGPT, partly Claude) and an independent-RIA lean (Perplexity, Grok, partly Gemini).
The opportunity is not to dominate the category — that’s structurally hard for any single RIA at the $1B–$50B tier — but to be present in both universes. That means corroboration in both the Barron’s/FA-Magazine ranking apparatus (where ChatGPT and Claude weight heavily) and in the regional family-office press + RIA-specific trade publications (where Perplexity and Grok weight heavily). Local and regional RIAs are getting buried in this fragmentation: the engines surface national names with deep corroboration, while regional firms surface only on prompts narrowed to geographic context (which this pilot did not test).
Underneath, surfacing in this category appears most sensitive to peer-recognition corroboration (Barron’s, FA Magazine, Forbes, RIA Edge), then entity clarity (the engines stop fragmenting your firm across slight name variants), then retrieval surface (Form ADV linked, advisor profiles parseable), then answer-shaped content (passage-citable answer blocks on HNW scenarios). Where you sit in the trust-bank-vs-independent-RIA lean depends largely on which corroboration set your firm has invested in.
Local and regional RIAs are getting buried. The engines surface national names with deep corroboration; firms outside the Barron’s Top 100 / FA Magazine apparatus and outside the family-office trade press largely do not surface on unbranded HNW buyer prompts. Compliance note: observed AI surfacing, not adviser quality, not investment advice, not solicitation.
Two shortlist universes — trust-bank vs. independent-RIA
Research publication based on sampled AI outputs collected on 2026-05-30. Findings reflect observed outputs in this sample and are not statements of company quality, ranking factors, or business performance.