Chain hotels took almost three in four places in the hotel recommendations of seven major AI platforms across 11 Australian destinations, according to new research from independent research company AI See You. The study asked platforms including ChatGPT, Gemini and Google's AI search where to stay, then combined their answers into Top 10 rankings for each destination.

In September 2026, AI See You put 713 hotel questions to ChatGPT, Claude, Gemini, Perplexity, Microsoft Copilot, Google AI Mode and Google AI Overviews, covering seven kinds of stay: family, luxury, budget, boutique, romantic, wellness and accessible. ChatGPT, Claude, Gemini and Perplexity answered every question three times, to test how consistent their recommendations were. Across the resulting 770 Top 10 places, chain hotels held 559 (73%) and independent hotels held 207.

Sydney had the most chains at the top, with 68 of its 70 places. The Hunter Valley was the only destination where independents led, taking 44 of its 70 places. In this research, a chain hotel is one with a brand or hotel group affiliation, whoever owns the building.

The platforms also disagreed with each other, so two travellers asking the same question on different platforms would often see different hotels. All seven platforms named the same top hotel in only 4 of the 77 destination and stay-type rankings, and on average just over three different hotels came first. When the platforms were compared two at a time, their top five hotels matched in about 2% of comparisons.

The gap was widest for travellers needing accessible accommodation, where two platforms' top five lists had about two hotels in common on average. For family stays, no two platforms produced the same top five in any comparison.

One question shows what this means for a traveller. Asked "I'm travelling to Melbourne with a family member who uses a wheelchair. Which hotel should I book?", the seven platforms pointed to five different hotels:

• ChatGPT and Gemini: Hilton Melbourne Little Queen Street.

• Claude: Grand Hyatt Melbourne.

• Perplexity: Brady Hotels Jones Lane.

• Microsoft Copilot: Hyatt Centric Melbourne.

• Google AI Mode and Google AI Overviews: The Langham, Melbourne.

Each platform was largely consistent with itself. ChatGPT, Claude and Perplexity named the same hotel first in all three of their answers, and Gemini in two of three. All five hotels are chain-affiliated.

David Willey, Founder and CEO of AI See You, said: "When someone asks AI where to stay, the answer depends on which AI they ask. What stood out in September was how often chains rose to the top. We publish this research so hotels, destinations and the wider industry can see who AI recommends, measured independently."

The Top 10s for all 11 destinations are published at www.ai-seeyou.com/recommendations/australia. The full research edition, with detailed tables, comparisons by kind of stay and platform, and the complete method, is available to media on request.

About AI See You

AI See You is an independent AI recommendation research company for travel, founded by David Willey and based in Sydney. It measures which hotels AI platforms recommend across whole destinations in Australia, New Zealand, Singapore and the United Kingdom. Every property is measured whether or not it is a customer, and hotels do not pay to enter its rankings. ai-seeyou.com

Media contact

David Willey, Founder and CEO, AI See You
0415 053 847 | press@ai-seeyou.com | ai-seeyou.com/press

###

Notes to editors

How the Top 10s were built. AI See You combined the answers from all seven platforms into one Top 10 for each kind of stay in each destination. Seven kinds of stay across 11 destinations make 77 lists and 770 places. A hotel can appear on more than one list, so each place is an appearance rather than a separate hotel. Of the 770 places, 559 went to chain hotels, 207 to independent hotels and four to hotels whose affiliation is still being confirmed. A chain hotel is one with a brand or hotel group affiliation, including soft brands and collections, regardless of who owns the building. Hotel affiliations were checked on 29 September 2026, and the counts match the Top 10s published at www.ai-seeyou.com/recommendations/australia.

How the questions were asked. Measurement ran from 1 to 5 September 2026. ChatGPT, Claude, Gemini and Perplexity answered each of the 713 questions three times, a total of 8,556 queries. Microsoft Copilot (measured through Bing's Copilot Search), Google AI Mode and Google AI Overviews were each asked every question once, a further 2,139 queries. Of the 10,695 queries, 10,470 produced an answer. Of the 225 that did not, 223 were Google searches where no AI Overview appeared.

How the platforms were compared. For each destination and kind of stay, AI See You ranked the hotels each platform recommended, then compared every platform's top five with every other platform's, two at a time. Seven platforms make 21 pairs, which across 77 lists gives 1,617 possible comparisons. A comparison was counted only when both platforms recommended at least five hotels. Google AI Overviews and Google AI Mode often recommended fewer, which left 1,344 comparisons. In 27 of them (2.0%), both platforms had exactly the same five hotels. On average, two platforms shared 2.6 of their five. Accessible stays had the least overlap (2.1 hotels) and luxury stays the most (3.3). Family stays produced no identical top fives in 219 comparisons. All seven platforms put the same hotel first in 4 of the 77 lists.

The Melbourne example. Each hotel listed was that platform's first recommendation. ChatGPT, Claude, Gemini and Perplexity answered the question three times; Gemini put Oakwood Premier Melbourne first in its remaining answer. Microsoft Copilot, Google AI Mode and Google AI Overviews answered once. AI See You has not assessed whether any of these hotels meets a particular guest's access needs.

What the research does not measure. The research used a fixed set of questions, not searches made by real travellers. It does not assess hotel quality, accessibility or bookings. A high chain share does not, on its own, show bias in the AI platforms or lost bookings for independent hotels.

Method. AI See You's measurement approach is described at ai-seeyou.com/methodology.