How often an insurance brand gets mentioned by AI chatbots and how often it actually gets recommended are two distinct and sometimes contradictory measures, according to a new benchmark from Brainpan.AI.
The Q2 2026 Insurance AIVI Benchmark evaluated 300 consumer-style insurance-shopping prompts across ChatGPT, Gemini, Perplexity, Copilot and Claude, generating 1,500 responses containing 4,328 organic mentions of 134 insurance brands.
The study found no single insurer leads across every category measured. GEICO led in auto insurance, State Farm led in home and renters, MassMutual led in life, and The Hartford led in commercial, with category rankings based on a combined measure of position, recommendation and citation performance.
The study's most notable finding centers on a gap between appearing in an AI response and actually being recommended within it.
The Hartford recorded an 88.1% top-three placement rate, the highest among the study's ten highest-ranked brands, yet only 27.6% of its appearances included an active recommendation, the lowest recommendation rate in that same group. Kevin Walsh, Brainpan.AI's founder, said that divergence was the study's central takeaway.
"The most important pattern wasn't simply which brands appeared. It was how often retrieval and recommendation diverged. A carrier can be highly visible and still lose the recommendation at the final step. All this in a channel many companies are not yet measuring," Walsh said.
The study also found market share offers little predictive value for AI visibility. In home insurance prompts specifically, Amica's share of AI-generated visibility ran 10.8 times higher than its NAIC-reported market share, an example the study used to argue that traditional measures of brand scale and market performance don't reliably translate into how often AI systems surface a given carrier.
"Being visible in an AI response and winning its recommendation are not the same thing. Some carriers appeared frequently but converted few of those appearances into recommendations. We also found smaller brands whose share of AI visibility substantially exceeded their NAIC-reported market share," Walsh said.
Brainpan.AI's benchmark arrives alongside a small but growing set of similar efforts to quantify how AI systems represent brands. Marketing research firm Conductor has separately published its own 2026 Insurance AI Search & AI Overviews Benchmarks examining which insurance brands AI cites most frequently and what content earns that placement, while insurance-specific AI maturity tracking has also emerged through Evident Insights' AI Index for Insurance, though that index measures carriers' own internal AI capability and investment rather than how AI systems represent those carriers to consumers, a meaningfully different question from the one Brainpan.AI's benchmark addresses.
For insurance marketing and brand teams, the study's core implication is that AI visibility now functions as a distinct channel requiring its own measurement, separate from traditional search engine optimization or paid placement strategy, since a carrier's search rankings or ad spend don't appear to translate predictably into how often or how favorably AI systems surface that brand to a consumer asking for coverage recommendations.
The Hartford's example is a useful illustration for any carrier reviewing its own digital marketing strategy: strong visibility alone doesn't guarantee conversion at the recommendation stage, meaning marketing teams evaluating AI-channel performance should track recommendation rate as a distinct metric from raw mention frequency rather than assuming the two move together.
