AI Index: Australian Challenger Lenders Beat Big Banks on Trust
Third Hemisphere's first AI index shows ING and OurMoneyMarket outscoring NAB, Westpac and ANZ on Reputation, with CommBank leading overall in Australia's AUD $168 billion market.

Why it matters
- ING and OurMoneyMarket scored 69/100 on Reputation, ahead of NAB (60), Westpac (56) and ANZ (42).
- The index analysed 166 unbranded questions across ChatGPT, Claude, Gemini, Copilot and Perplexity, covering Australia's AUD $168 billion personal lending market.
- CommBank posted the highest Reputation (71) and Presence (46) scores, with an overall CRISP score of 54 tied with NAB for first place.
- 28 lenders made the leaderboard; 96 were classed as under the radar for appearing in fewer than 30 answers.
- Third Hemisphere said 87% of Australians now use AI search for financial services purchasing decisions.
ING and OurMoneyMarket scored 69 out of 100 on Third Hemisphere's new Reputation measure for AI-generated lending advice, beating NAB on 60, Westpac on 56 and ANZ on 42. The figures come from the first edition of an index that measures how AI engines portray Australian personal lenders, launched by Third Hemisphere and covering the country's AUD $168 billion personal lending market.
The index assesses how lenders are recommended, cautioned against and cited by AI search tools and chatbots. It groups lenders by how often engines mention them and whether those mentions are favourable or negative. The core finding inverts conventional market hierarchy: smaller lenders can perform as well as, or better than, larger banks on recommendation strength, even when they appear less often in AI-generated answers.
Plenti scored 68 on Reputation and Harmoney 67, both ahead of the major banks except one. CommBank recorded the highest Reputation score at 71, along with the highest Presence score at 46, meaning engines named it in 46% of relevant answers under the methodology. Its overall CRISP score of 54 matched NAB for the top ranking.
What is CRISP and how was it built?
The ranking draws on 166 unbranded questions covering personal loans, debt consolidation, car finance, credit cards and short-term credit. The questions span five stages of a borrower's decision-making process: learning, finding, comparing, choosing and buying. Of the 166 questions, 85 asked what to choose and 81 asked what to avoid. None named a lender directly.
For the lending study, Third Hemisphere used answers from five AI engines: ChatGPT, Claude, Gemini, Copilot and Perplexity. The company said its wider dataset includes 195,689 AI answers across 159 sectors and seven AI engines, generating 1,526,848 brand mentions and identifying 60,206 distinct companies.
The scoring system, known as CRISP, uses five measures: Consensus, Reputation, Influence, Source and Presence. Reputation tracks whether engines recommend a company more often than they warn against it. Presence measures the share of answers in which the company is named.
Scores are assigned only where a lender appears in at least 30 answers. On that basis, 28 lenders made the leaderboard, while 96 were classed as under the radar because they were named too rarely for a stable score.
Who won and who lagged?
The index groups lenders into six bands on what it calls a Star Chart:
- AI Famous — mentioned frequently and recommended more often than cautioned against: CommBank, NAB and Westpac.
- Rising Star — ING and Macquarie.
- Cult Favourite — 15 lenders that appeared less often but received favourable treatment when they did, including Plenti, OurMoneyMarket, Harmoney, NOW Finance and Bendigo Bank.
- AI Infamous — mentioned often but cautioned against more than recommended: ANZ, the only lender placed in this category.
- Off the Guest List — mentioned relatively rarely with more cautionary than supportive references: Cash Converters, BOQ, Latitude Financial Services, St.George, Nimble and Resimac.
Several well-known consumer finance brands barely featured in the sample. Afterpay, Zip, Klarna, Revolut and Virgin Money were each named in fewer than 30 answers, leaving them outside the scored group.
The research also found that comparison sites including Finder and Canstar carried significant influence at the point where borrowers choose a lender. That reflects the role of third-party sources in shaping how AI engines respond to finance-related prompts.
Why does this matter for the lending market?
The study lands as consumers increasingly consult AI tools before visiting provider websites. Third Hemisphere said 87% of Australians now use AI search for financial services purchasing decisions. For banks and challengers competing in a AUD $168 billion market, the composition of an AI answer can shape a borrower's shortlist before any direct contact with a brand.
Jeremy Liddle, Executive Chair of Third Hemisphere, said: "The good news for every lender is that those sources can be influenced. The engines form their opinions from comparison sites, ranking sites, regulators and the media, so companies that earn real coverage, engage those sources well and publish facts people can rely on can earn their way into the answers."
Liddle framed the stakes in direct terms. "AI engines now sit between Australians and the decisions they make about their money, and when someone asks ChatGPT which lender to trust, the answer shapes their shortlist before they visit a comparison site or a bank's website," he said. "Authentic leadership is the future of marketing. I built the CRISP technology so Third Hemisphere could help Australian brands lead the world, and so every Australian company can see how the AI engines describe it, with the tools and support to improve its standing."
What should lenders do next?
Hannah Moreno, Founder and CEO of Third Hemisphere, connected the index to the company's longer history in financial services reputation work. "Trust has been the currency of financial services long before I founded Third Hemisphere in 2014 with MONEYME as our first client," Moreno said. "What is different now is the AI engines decide who deserve this trust as guardians of the information we consume about brands. Third Hemisphere has spent more than a decade building that trust for the companies we represent across multiple channels, so evolving to measure it in AI answers was a natural next step."
Moreno said the first index gives every featured lender a specific task. "Challengers already win the engines' trust when they're found, so their task is to be found more often, while three of the big four need to change what the engines say about them," she said. "Both are fixable, and we're happy to show any lender how."
The split is visible in the numbers. Challengers such as ING, OurMoneyMarket, Plenti and Harmoney already carry Reputation scores in the high 60s, but most sit in the Cult Favourite band with limited Presence. ANZ, by contrast, carries the big-four profile of frequent mentions combined with the lowest Reputation score among the majors at 42.
The index implies two distinct work programmes for Australian lenders: earn more mentions, or change what the engines say. Both routes run through the same intermediary sources — comparison sites, ranking sites, regulators and the media — that currently shape AI answers on personal lending.
Source: itbrief.com.au
More from Marcus Bennett
Show full bio
Senior reporter covering companies, deals and strategy in artificial intelligence and machine learning at AI In Context.
618 articles
Related articles
- AI Assistants Now Write the B2B Shortlist — and 60% of the Lists They Trust Are Competitors' Own Posts
- Community banks are deliberately slow-walking AI adoption
- Australia Warns AI Will Industrialise Scams as Losses Hit $2.18 Billion
- Only 30% of Australians Think AI's Benefits Outweigh Its Risks
- Bank AI Adoption Stalls on Human Fears, Not Technology