AI search concentrates insurance brands in Australia, Somantra says

Somantra says AI search is funneling Australian insurance visibility toward a small group of brands while leaving most long-tail questions unanswered. The report also finds Google AI Overviews and ChatGPT often recommend different brands, creating a widening opening for insurers that can win structured AI citations.

Why it matters: - AI search is becoming a new gatekeeper for insurance discovery in Australia. - Somantra says a small group of brands is capturing most branded mentions, while 70% of detailed insurance questions still produce no brand recommendation. - That split creates both concentration risk for smaller insurers and a clear opening for brands that can win citations in AI answers.

What happened: - Somantra released its May 2026 AI Search Visibility Report covering 20 Australian insurance brands across 34,278 real consumer conversations on Google AI Overviews and ChatGPT during May 2026. - The report found that on ChatGPT, Allianz, NRMA and AAMI accounted for 50% of all insurance mentions. - Five brands accounted for 60% of ChatGPT mentions, seven accounted for 75%, and nine accounted for 90%. - On Google AI Overviews, four brands made up 50% of mentions, five reached 60%, eight reached 75%, and eleven covered 90%. - Somantra founder Arun Prasad said Google gives users options, while ChatGPT gives a shorter shortlist that is getting shorter.

The details: - The report found concentration at the product level as well. - Allianz held 3,941 mentions in travel insurance. - NRMA led car insurance with 3,238 mentions. - QBE dominated motorcycle insurance with 2,896 mentions. - Budget Direct led pet insurance with 940 mentions. - In several categories, the next-closest competitor trailed by a wide margin. - The report also found that 70% of long-tail AI-generated answers named no insurance brand at all. - Examples of those brandless questions included how to claim pet insurance for an overseas trip and which discounts apply when insuring a new car. - Somantra says the same forces that concentrate generic queries at the top also leave specific questions open because AI engines have not settled on default brands for them. - More information is available in the May 2026 report.

Between the lines: - The data suggests AI search is not functioning like traditional search, where many results can compete for attention. - Instead, each platform appears to be building its own default shortlist of trusted brands. - That matters because an insurer can lose visibility on one platform but still have an opening on the other. - The report found Google AI Overviews and ChatGPT agreed on the leading recommended brand for the same query only 27.9% of the time, up from 23.7% in March. - Even when both platforms named a brand, they chose different brands about seven times out of ten. - Somantra argues that divergence widens the opportunity surface for brands that can tailor content to the sources AI engines trust.

What's next: - Somantra says the race is moving quickly, not slowly. - The number of unique domains AI engines cited as trusted sources fell 21% from March to May, dropping from 10,777 to 8,488. - Prasad said the window is closing and that brands waiting for AI search to mature risk ceding ground to competitors. - The company expects more long-tail questions to be claimed by brands that publish structured content on sources AI engines already use.