US Insurance AI Discovery Index — Edition 1

Who does AI recommend when buyers ask about insurance?

We measured 88 US insurance organizations across ChatGPT, Perplexity, Gemini, and Claude on name-free buyer questions — every answer stored in a tamper-evident evidence chain.

US Insurance AI Discovery Index — Edition 1 · June 2026 · 88 organizations · methodology 2026.1

65%

of the 88 measured organizations (57) appear in no buyer answer on Perplexity.

49%

are invisible on ChatGPT (43 of 88) — while Google Gemini names most of the same organizations.

88

organizations in the frozen Edition-1 cohort (June 2026) — segments are never ranked against each other.

The finding: visibility is engine-specific

The same cohort, the same questions — four very different answers.

AssistantInvisible (of 88)Avg. visibilityRegional P&C invisibleHealth plans invisible
ChatGPT43 (49%)15%33/38 (87%)10/50 (20%)
Claude43 (49%)16%17/38 (45%)26/50 (52%)
Google Gemini4 (5%)82%3/38 (8%)1/50 (2%)
Perplexity57 (65%)9%38/38 (100%)19/50 (38%)

Combined across all four assistants, only 2 of 88 organizations (2%) are invisible everywhere — an organization counts as visible on a question when at least one assistant names it. That average hides the story: the assistant your buyer happens to use decides whether you exist in the answer.

Regional health plans

Each dot is one regional health plan in the Edition-1 cohort. 0 of 50 (0%) never appear in any name-free buyer answer on any of the four measured assistants — per-assistant gaps are far larger (see the table above). The invisible are reported only as this anonymous count.

Who the AI named — leaders by presence

  • AmeriHealth100% of questions
  • AvMed100% of questions
  • Baylor Scott & White Health Plan100% of questions
  • Blue Cross and Blue Shield of Alabama100% of questions
  • Blue Shield of California100% of questions
  • BlueCross BlueShield of South Carolina100% of questions
  • BlueCross BlueShield of Tennessee100% of questions
  • CDPHP100% of questions

Names appear because the assistants named them — frequency is a measured fact, not our rating or endorsement.

Regional P&C carriers

Each dot is one regional p&c carrier in the Edition-1 cohort. 2 of 38 (5%) never appear in any name-free buyer answer on any of the four measured assistants — per-assistant gaps are far larger (see the table above). The invisible are reported only as this anonymous count.

Who the AI named — leaders by presence

  • Acuity Insurance100% of questions
  • Auto-Owners Insurance100% of questions
  • Central Insurance100% of questions
  • Cincinnati Insurance100% of questions
  • Concord Group Insurance100% of questions
  • Donegal Insurance Group100% of questions
  • EMC Insurance100% of questions
  • Encova Insurance100% of questions

Names appear because the assistants named them — frequency is a measured fact, not our rating or endorsement.

What an answer actually looks like

A buyer asks

best business insurance provider for mid-size businesses

AI assistants name
The HartfordTravelersNationwide

Evidence receipts: pub-fe0d55d022d3 · pub-b75398e8fa09

A buyer asks

best health insurance plan for mid-size employers

AI assistants name
UnitedHealthcareAetna

Evidence receipts: pub-ceaf310c4981 · pub-d131ede2e6a3

Cohort and method

Inclusion criteria: US regional and mid-market insurance carriers (50 health, 38 commercial P&C) with an owned website and active digital presence, selected before measurement in June 2026. National giants are tracked only as competitor mentions, not ranked members. Cohort frozen 2026-06-01; capture period June 2026; methodology 2026.1 (full methodology). Every published number is regenerated from the frozen evidence and verified before publication — nothing is hand-copied.

Disclosures

  • Benchmark metric computed on: chatgpt, claude, gemini, perplexity.
  • Headline visibility counts a question as covered when at least one measured engine names the organization; per-engine distributions are reported separately because engines differ substantially.
  • Canonical discovery questions: p1, p10, p12, p14, p4, p7.
  • Capture window: 2026-06-01 through 2026-06-30 (inclusive).
  • 419 answer(s) captured before methodology stamping are attributed to methodology 2026.1 by operator assertion.
  • This edition has no sponsors. Sponsorship never buys a ranking, suppresses a result, or changes methodology.

Corrections and right of reply

No corrections have been issued for this edition. Measured organizations may request the evidence behind their result or submit a correction at info@nextaiforge.com. Corrections never mutate the original evidence; they are logged publicly here. Organization-level negative detail is shared privately with that organization only.

Want your number — and why?

The index shows the market. The AI Visibility Report shows your brand: the same four-assistant measurement, question by question, with every answer preserved verbatim — delivered in 48 hours for a fixed $290.