Which home loan lenders does AI recommend in Australia?
We asked ChatGPT and Claude to recommend home loan lenders across 8 Australian questions, registered its checks before looking, and found the mid-tier lenders holding a Canstar award — Macquarie and Unloan — appear at 2.4× the rate of those without.
Australia's AI home-loan shortlist is not a big-banks-only echo. The two mid-tier lenders holding a current Canstar award — Macquarie and Unloan — appear at a 28.7% recommendation rate, 2.4× the 12.1% averaged by the four mid-tier lenders without one. Both fictional control lenders scored 0%.
Every tracked lender appears in a Google search. AI names fewer — and the mid-tier brands it surfaces most often hold a Canstar award.
When you Google a home loan in Australia, the big four banks take the page. A mid-tier lender spending on awareness — Macquarie, ING, Athena, Unloan, UBank, Reduce — can buy recognition and still not appear when a shopper asks ChatGPT or Claude for a recommendation. The brands investing in being known are not always the brands an AI names. The question is which ones it does name, and why. A lender counts as recommended on a call when the engine names it in its answer; its rate is the share of calls that name it.
At the top, AI echoes market position
Commonwealth Bank leads the eight tracked lenders on both engines — 42.2% of calls on ChatGPT and 64.5% on Claude — and it leads the consumer funnel too: a 2025 Conjointly survey puts Commonwealth Bank at 17.9% preference share, more than three times the nearest rival, with 52.5% awareness and 32.3% consideration. Westpac, the second major bank, sits at 37.0% on ChatGPT and 36.8% on Claude. For the two big banks, AI recommendation is directionally consistent with the Conjointly survey — CommBank leads both.
Below CommBank, the shortlist follows a Canstar award
Each lender’s share of API calls on which it was named, across 8 questions × 17 repeats × 3 batches. Bars above the 15% floor are in-band. ★ marks lenders holding a cited Canstar award.
The assumption that AI echoes market share breaks at the next tier down. Of the six mid-tier lenders we tracked, the two holding a current Canstar Outstanding Value or Bank of the Year award — Macquarie and Unloan — averaged a 28.7% recommendation rate across both engines. The four without one — ING, Athena, UBank, Reduce — averaged 12.1%. That is a 2.4× gap. Macquarie sits in the band on both engines (35.0% ChatGPT, 36.0% Claude). Unloan sits in the band on Claude at 34.1% (9.6% on ChatGPT, below the band). Macquarie holds the 2025 Bank of the Year for fixed-rate home loans; Unloan holds the 2026 Canstar Outstanding Value Award — Variable Home Lender. The mid-tier brands with the highest recommendation rates are not random — they are the ones a comparison site has already singled out.
An award is a page the AI already has to retrieve
Average recommendation rate across both engines for the six mid-tier lenders, split by whether they hold a current Canstar Outstanding Value or Bank of the Year award. Individual brand averages shown beneath each group mean.
A Canstar award is not a signal the model reads off a trophy. It is a generator of pages. An award produces editorial explainers, ranked comparison tables, and linked lender profiles — the kind of documents a retrieval-based answer engine reads. The three AU comparison sites tell different stories. InfoChoice's home-loans page lists all eight tracked lenders, so it does not separate the ones an AI names from the rest. RateCity now redirects to canstar.com.au — its standalone leaderboards are gone. Canstar is the column that discriminates: of the mid-tier, only Macquarie and Unloan currently hold an award. In this data, the lenders an AI names are the ones a comparison site has already written up. We measured which lenders each engine named, not which pages it fetched, so whether those pages cause the naming is a question for a retrieval-logging study, not this one.
Claude named seven of eight; ChatGPT named three
Number of the 8 real lenders whose day-0 recommendation rate sat in the in-band window [0.15, 0.85] on each engine.
The two engines agree on CommBank, Westpac, and Macquarie — all three land in the band on both. After that, the lists diverge. Claude placed 7 of 8 real lenders in the band; ChatGPT placed 3. Unloan, Athena, Reduce, and ING all appeared on Claude’s shortlist but fell below the 15% floor on ChatGPT. Both fictional controls — Kestrel Home Loans and Veldra Finance — scored 0% on both engines, confirming neither engine hallucinated a lender.
Eight questions, 816 calls, one number locked before we looked
Eight phrasings, seventeen repeats each, three batches per engine, two engines. Each call scores all ten tracked brands.
- best home loans Australia
- top home loan lenders in Australia
- best variable rate home loan Australia
- lowest interest rate home loan Australia
- best home loan for first home buyers Australia
- best fixed rate home loan Australia
- recommend 5 home loan providers in Australia and say why
- best offset mortgage Australia
We asked both engines the same 8 questions a home-loan shopper would type. Each question ran across 8 real lenders (the two big-bank majors plus six mid-tier) and 2 fictional control lenders we invented as a quality check, 17 repeats each, on both ChatGPT and Claude with live web search, on day-0 only. That is 408 calls per lender per engine, 816 calls in total. A lender is in the band when its recommendation rate sits between 0.15 and 0.85. Before we read a single result we locked one go/no-go number: at least three mid-tier lenders in the band on one engine, or the study stops at a one-batch observation. Five of the six mid-tier lenders cleared that bar on at least one engine. Two registered checks confirmed the formula’s intervals behave as advertised (98.5% coverage, above the 90% threshold) and found no evidence that recommendation rates scatter more than chance across day-0 batches (observed 90.78, under the 117.24 chance-only high mark).
Day-0, six mid-tier brands, one country
The 2.4× Canstar gap is a descriptive on six mid-tier lenders, not a registered bet and not a cause — having an award and being recommended move together in this data, but six brands cannot separate correlation from coincidence. This is one day, one country, two engines. The one-week and two-week stability arms have not run, so this says nothing yet about whether the shortlist drifts. Eight brands is not the Australian market; lenders we did not track, including ANZ and NAB, are not scored. These queries ran against the OpenAI and Anthropic APIs with live web search, not the consumer ChatGPT or Claude products (the consumer-vs-API gap is detailed in the limitations below).
The pattern points to the comparison site, not the lender’s own site
A mid-tier lender without a comparison-site award has fewer stand-out pages for an AI to cite, and in this data it shows up at less than half the rate of the two that hold one. The shortlist an AI names follows the comparison-site layer that already existed: the lenders that comparison editors had already singled out are the lenders with the pages an AI retrieves, and the lenders with pages are the lenders it names. For a mid-tier brand, the pattern in this data is not about its own site; it tracks whether a comparison site has already written it up.
How we ran this
- What we measured
- 8 Australian-specific home-loan questions × 8 real lenders (2 major-bank majors + 6 mid-tier) + 2 fictional control lenders, 17 repeats each, on ChatGPT and Claude with live web search, day-0 only. A lender is in-band when its recommendation rate is between 0.15 and 0.85. The registered metric is per-lender recommendation rate pooled across the 8 questions, with Wilson 95% intervals labelled illustrative. Because one call can name several lenders, outcomes within a call are not independent, so these intervals are descriptive, not inferential. Australian home loans was the first category through the stability study's pilot gate — a high-consideration market with a defined mid-tier — so the choice was set before any rates were read.
- When
- Data collected 21–22 July 2026 (day-0 window). Analysis published 22 July 2026.
- Which AI
- ChatGPT (OpenAI API) with live web search, Claude (Anthropic API) with live web search, Gemini — not measured; its API hides fetched URLs, which contaminates a rate study
The full numbers
Every number in this piece, at full precision. The prose rounds for reading; this table doesn't.
| Measure | As shown | Exact | Source |
|---|---|---|---|
| Canstar-awarded mid-tier vs the rest — rate ratio | 2.4x | 2.3756 | Derived |
| Canstar-awarded mid-tier — mean recommendation rate | 28.7% | 28.6765% | Derived |
| Mid-tier without a Canstar award — mean rate | 12.1% | 12.0711% | Derived |
| Commonwealth Bank — recommendation rate, both engines | 53.3% | 53.3088% | Measured |
| Westpac — recommendation rate, both engines | 36.9% | 36.8873% | Measured |
| Commonwealth Bank — ChatGPT recommendation rate | 42.2% | 42.1569% | Measured |
| Commonwealth Bank — Claude recommendation rate | 64.5% | 64.4608% | Measured |
| Westpac — ChatGPT recommendation rate | 37.0% | 37.0098% | Measured |
| Westpac — Claude recommendation rate | 36.8% | 36.7647% | Measured |
| Macquarie — ChatGPT recommendation rate | 35.0% | 35.0490% | Measured |
| Macquarie — Claude recommendation rate | 36.0% | 36.0294% | Measured |
| Unloan — ChatGPT recommendation rate | 9.6% | 9.5588% | Measured |
| Unloan — Claude recommendation rate | 34.1% | 34.0686% | Measured |
| ING — ChatGPT recommendation rate | 12.25% | 12.2549% | Measured |
| ING — Claude recommendation rate | 16.42% | 16.4216% | Measured |
| Athena — ChatGPT recommendation rate | 2.21% | 2.2059% | Measured |
| Athena — Claude recommendation rate | 20.10% | 20.0980% | Measured |
| UBank — ChatGPT recommendation rate | 7.35% | 7.3529% | Measured |
| UBank — Claude recommendation rate | 8.58% | 8.5784% | Measured |
| Reduce Home Loans — ChatGPT recommendation rate | 11.76% | 11.7647% | Measured |
| Reduce Home Loans — Claude recommendation rate | 17.89% | 17.8922% | Measured |
| Share of intervals that held their rate | 98.5% | 98.4848% | Measured |
| Coverage threshold | 90% | 90.0000% | Measured |
| Batch-scatter statistic, observed | 90.78 | 90.7789 | Measured |
| Batch-scatter statistic, chance-only high mark | 117.24 | 117.2369 | Measured |
| Mid-tier lenders in-band on at least one engine | 5 | 5 | Measured |
| Mid-tier lenders tracked | 6 | 6 | Measured |
| Real lenders in-band on ChatGPT | 3 | 3 | Measured |
| Real lenders in-band on Claude | 7 | 7 | Measured |
| Fictional control lenders — recommendation rate | 0.0% | 0.0000% | Measured |
| Fictional control lenders — extracted at all | 0 | 0 | Measured |
| Day-0 calls (both engines) | 816 | 816 | Measured |
| Calls per lender per engine | 408 | 408 | Measured |
| Repeats per phrasing | 17 | 17 | Measured |
| Day-0 batches per engine | 3 | 3 | Measured |
| Question phrasings | 8 | 8 | Measured |
| Real lenders tracked | 8 | 8 | Measured |
| Fictional control lenders | 2 | 2 | Measured |
| AI systems measured | 2 | 2 | Measured |
| Day-0 measurement spend (US$) | 69.36 | 69.3600 | Measured |
| Phase A study ceiling (US$) | 160 | 160 | Measured |
| ChatGPT — cost per call (US$) | 0.065 | 0.06500 | Measured |
| Claude — cost per call (US$) | 0.105 | 0.10500 | Measured |
| Conjointly — Commonwealth Bank awareness | 52.5% | 52.5% | Cited |
| Conjointly — Commonwealth Bank consideration | 32.3% | 32.3% | Cited |
| Conjointly — Commonwealth Bank preference share | 17.9% | 17.9% | Cited |
What this doesn't settle
- One day, one country, two engines — the rates are a finding for AU home loans on the OpenAI and Anthropic APIs on these dates, not a number that generalises across time, engines, or geographies.
- Association is not cause — the 2.4× Canstar gap is a descriptive on six mid-tier lenders, not a registered bet and not a cause. Having an award and being recommended move together in this data; six brands cannot separate correlation from coincidence.
- Day-0 only — the one-week and two-week stability arms have not run, so this says nothing yet about whether the shortlist drifts over a fortnight.
- Eight brands is not the market — lenders not tracked, including ANZ and NAB, are not scored. The study answers which of the tracked lenders AI names, not which lenders.
- These queries ran against the OpenAI and Anthropic APIs with live web search — not the consumer ChatGPT or Claude products. The APIs carry no personalisation, no conversation memory, and no geo-targeting.
References
- Mortgage Provider Awareness, Consideration & Preference – Australia Accessed 22 July 2026.
- Canstar Bank of the Year – Fixed Rate Home Loans Award Accessed 22 July 2026.
- Canstar Home Loans Star Ratings & Awards Accessed 22 July 2026.
- Home Loans — Compare Home Loan Rates Accessed 22 July 2026.
- RateCity Home Loans (now redirecting to Canstar) Accessed 22 July 2026.
Is your brand on the comparison pages an AI reads before it names a lender?
A diagnosis maps which pages each AI system reads before it recommends in your category — and whether your brand is on them. Not a score: the evidence, run twice.