The engines tell consumers Mayo wins Minnesota, Arizona, and Florida — and surface Mayo as a benchmark across twenty-three additional states where no Mayo facility exists.
HCA wins zero.
The 5W AI Trust Map: How the Finding Was Measured
The 5W AI Trust Map of America — Healthcare Volume, May 2026, scored fifty states across five AI engines (ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews) on twelve buyer-intent prompts per state. Three thousand data points. Acute care ("best hospital for emergency care in [state]"). Specialty care ("best hospital for cancer treatment in [state]"). Consumer trust ("most trusted hospital system in [state]").
The pattern is dominant: 46 of 50 states cite an academic medical center first.
The Academic Medical Center Dominance Pattern
Mayo dominates Minnesota and ranks second in 14 additional states. Cleveland Clinic dominates Ohio and ranks second on cardiology in 11 additional states. Johns Hopkins dominates Maryland and ranks second on research medicine in 9 additional states. Mass General dominates Massachusetts. NYU Langone dominates New York. MD Anderson dominates Houston. Penn Medicine dominates Pennsylvania. Cedars-Sinai dominates Los Angeles. UCLA Health dominates Western California state-level prompts.
The dominant institutions share four structural features the engines weight heavily.
The Four Signals AI Engines Actually Weight
Peer-reviewed research output. Mayo published over 9,000 peer-reviewed papers in 2024. Cleveland Clinic, over 5,000. Johns Hopkins, over 12,000. The engines retrieve the academic corpus as primary source material. Systems without comparable output do not generate the source density.
US News & World Report dominance. Mayo has held #1 on the Best Hospitals Honor Roll for nine straight years. Cleveland Clinic, a top-five position for two decades. Johns Hopkins, continuous Honor Roll placement since the ranking's inception. The engines treat US News as authoritative and pull from it heavily.
Founding-region editorial archive. Each dominant institution carries a multi-decade primary-source press archive in its founding region — Mayo in Rochester, Cleveland Clinic in Cleveland, Johns Hopkins in Baltimore. Local and regional coverage compounds across decades into an authoritative corpus.
Philanthropy reporting. Each runs a substantial foundation with IRS Form 990 filings, sustained philanthropy press, and named-gift coverage. The Mayo Clinic Foundation, the Cleveland Clinic Foundation, the Johns Hopkins Medicine Foundation — each generates primary-source content the engines retrieve.
The HCA Pattern: Zero-State Dominance Is Not a Measurement Artifact
HCA's zero-state finding held across prompt variants and session windows. HCA appears in engine answers when prompted by name. HCA does not appear as the engines' recommendation when prompted by need.
The four signals are largely absent from HCA's corpus relative to the academic peers in its operating markets. Peer-reviewed output is limited. US News Honor Roll placement is materially lower. The founding-region archive is anchored to corporate-operations coverage, not care-experience coverage. Philanthropy reporting is operational, not institutional.
The same finding holds for Tenet, CommonSpirit, Ascension, and the other large hospital chains. Each operates extensively. None dominate the AI answer in any state.
The Implication for Hospital Marketers
The Trust Map points to a category of communications infrastructure that diverges sharply from how most hospital systems currently practice marketing. Hospital marketing has historically optimized for service-line awareness, patient-acquisition campaigns, payer-relations narratives, and physician-recruitment positioning. None of these matter to the AI engines retrieving healthcare recommendations.
The signals that matter — research output, ranking placement, founding-region editorial density, philanthropy reporting — sit outside the marketing function. They sit inside research operations, philanthropy, and academic affairs.
The hospital systems that dominate the AI answer have, accidentally or deliberately, built primary-source infrastructure across functions other systems would not classify as marketing.
What the C-Suite Should Do About It
A hospital system's AI authority is now the function of investment categories that do not report to the CMO. Restoring the AI answer requires either reorganizing the comms function to include those categories or building genuine peer-reviewed, ranking-eligible, philanthropy-grade infrastructure inside the marketing function itself.
The largest US hospital chains are losing the AI answer across their operating footprints. The remediation is not a campaign. It is a multi-year investment in the primary-source layer the engines retrieve from.
For health-system C-suites: more than a third of US healthcare consumers now start care research inside AI engines. The AI answer is the first frame of patient consideration in markets where consumer choice drives care selection — elective procedures, cancer, cardiology, orthopedics, fertility, behavioral health. See EPR's coverage of patient discovery in aesthetic medicine for the same pattern at practice level.
The Through-Line
Academic medical centers built their primary-source corpora over six decades. They built them for research, ranking, and philanthropy. The engines now retrieve from those corpora as authoritative. The non-academic chains in their markets did not build comparable corpora. The asymmetry shows up in every state-level AI healthcare answer EPR has measured.
Mayo wins every state it touches. The structural reason is now documented. The competitive response is the next decade of hospital-system communications budget.
