---
title: "How lending platforms reach borrowers who ask AI about loans"
description: "By being named, with accurate rates, terms and reputation, when borrowers ask AI where to get a loan. Visibility brings applications, never approvals."
canonical: "https://underneath.agency/resources/lending-platforms-borrowers-ai-search"
published: 2026-10-07
updated: 2026-10-08
publisher: "Underneath (https://underneath.agency/agent)"
entity: "https://underneath.agency/.well-known/entity.json"
---
Guide · AI search

# How do lending platforms reach borrowers who ask AI where to get a loan?

By making sure AI assistants describe their loans, costs, eligibility and reputation accurately when people and small businesses ask where to borrow. Online lenders are taking a growing share of personal and small-business lending, and the first comparison often happens in a conversation, not on a rate table. Visibility can bring qualified applications; it can never promise anyone a loan.

## The short version

1. Online lenders lead personal-loan growth: [TransUnion](https://media.transunion.com/content/dam/transunion/us/business/collateral/report/1-fs/q2-2026-ciir-consumer-lending-report.pdf) reports fintech lenders’ share of unsecured personal loan originations rose from 39.8% to 45.0% in a year, with balances at a record $281 billion.
2. Small businesses are moving online too: in the Federal Reserve Banks’ [Small Business Credit Survey](https://www.fedsmallbusiness.org/-/media/project/clevelandfedtenant/fsbsite/reports/2026/2026-report-on-employer-firms/2026-report-on-employer-firms.pdf), the share of applicants that sought financing at online lenders rose from 17% to 29% over five years.
3. Cost surprises are the trust gap: 60% of firms that borrowed from online lenders said costs were higher than expected, against 37% at small banks and 32% at large banks.
4. Approval is never certain: the [Federal Reserve’s household survey](https://www.federalreserve.gov/publications/files/2024-report-economic-well-being-us-households-202505.pdf) found one-third of 2024 credit applicants were denied or approved for less than they requested.
5. Answers depend on the market: in [our country study](https://underneath.agency/research/ai-recommendations-by-country-study), questions such as lending, insurance and tax showed the largest country effect on the brands ChatGPT named.

A note before you read: this article is about how lenders appear in AI answers. Nothing here is financial, legal or compliance advice.

## Who borrows from lending platforms, and what is a funded loan worth?

Consumers consolidating debt or funding a purchase, and small businesses seeking working capital; value comes from funded loans, not clicks.

Lending platforms serve two very different borrowers. Consumers take unsecured personal loans, often to pay down credit cards. TransUnion counts personal loan balances at $281 billion, up 9.6% in a year, and originations up 19.5%. Fintech lenders drove that growth, with their originations up 35.0%, while banks’ share fell from 13.1% to 10.8%. Subprime borrowers made up about 38% of new loans, a reminder that many of these customers are stretched. Home loans follow their own path, covered in [how mortgage lenders win borrowers through AI](https://underneath.agency/resources/mortgage-platforms-leads-ai-search).

Small businesses borrow for cash flow, equipment and expansion. The [Small Business Credit Survey](https://www.fedsmallbusiness.org/-/media/project/clevelandfedtenant/fsbsite/reports/2026/2026-report-on-employer-firms/2026-report-on-employer-firms.pdf), with 6,525 responses from employer firms, found that large banks remain the top place firms apply, but that applicants at online lenders, such as OnDeck in the survey’s own examples, rose to 29%. Those applicants prioritized speed and their expected chance of being funded, while bank applicants chose on existing relationships. For the payments side of small-business finance, see [how payment processors reach merchants through AI](https://underneath.agency/resources/payment-processors-merchant-demand-ai-search).

Comparison marketplaces such as LendingTree, NerdWallet and Credit Karma sit between the two: they earn from referring borrowers to lenders. For a lender, the commercial value of a borrower is a funded loan that performs; for a marketplace, it is a qualified referral.

## Where does AI already sit in a borrower’s decision?

In early research, where many consumers and small businesses already use AI, though usually as a starting point.

An [Intuit Credit Karma survey](https://stacker.com/stories/personal-finance-investing/rise-fin-ai-why-americans-are-trusting-generative-ai-their) found that 66% of Americans who had used generative AI had used it to seek financial advice, and 85% of those had acted on what it said. The same survey adds a warning: 52% of those who acted said they had made a poor financial decision or mistake based on it, and 80% said they still research and validate the advice first. Credit Karma runs a lending marketplace, so read this as vendor research.

Small businesses use AI widely too. The Small Business Credit Survey found 46% of firms said the business or its employees currently use AI, most often for writing or marketing. We found no public survey measuring how many borrowers ask an AI assistant which lender to choose.

The Federal Reserve’s household survey frames the stakes. Thirty-four percent of adults applied for some type of credit in 2024, and twenty-one percent reported experiencing financial fraud or scams involving their money. Borrowers asking AI “is this lender legit?” have good reason to ask.

## What do borrowers ask AI before applying for a loan?

What they can get, what it will cost, whether a lender is trustworthy and whether checking will hurt their credit. We wrote the borrower prompts in the table ourselves as examples; none comes from a real applicant.

| Need | Illustrative prompt |
|---|---|
| Debt consolidation | “Best way to consolidate $15,000 of credit card debt with a 680 credit score” |
| Cost | “What APR should I expect on a personal loan with fair credit?” |
| Process | “Does checking my rate with an online lender hurt my credit?” |
| Trust | “Is [lender] legit, and what do borrowers complain about?” |
| Small business | “Online lender or bank for a $50,000 line of credit?” |
| Speed | “How fast can a restaurant get a working capital loan without collateral?” |
| Marketplace | “Is it better to use a loan comparison site or go straight to a lender?” |

The small-business questions track the survey closely. Online-lender applicants said speed and the chance of being funded drove their choice, and high interest rates and unfavorable repayment terms were their most common complaints. An answer that explains speed without cost serves the borrower badly and, we infer, sets up the cost surprise the survey found.

## How does a borrower get from an AI answer to a funded loan?

Through a rate check and an application; the lender’s underwriting decides, so visibility shapes who applies, not approval.

The path, as we infer it from how online lending works:

1. **Question.** “Which lenders consolidate card debt for fair credit?”
2. **Answer.** The assistant names lenders or marketplaces and describes rates and requirements.
3. **Rate check.** The borrower checks a rate, often with a soft credit inquiry, on the lender’s site or a marketplace.
4. **Application.** Documents, verification and a credit decision.
5. **Decision.** Approved, approved for less, or declined; the household survey shows one-third of applicants get the second or third outcome.
6. **Funded loan.** Revenue for the lender over the loan’s life; a referral fee for a marketplace.

The useful measure is therefore not traffic but qualified applications: borrowers who arrive understanding the product, its cost and its eligibility. An AI answer that oversells approval odds or understates rates sends applicants who will be declined or disappointed. Our guide to [linking AI answers to pipeline and revenue](https://underneath.agency/resources/ai-answers-pipeline-revenue) explains measurement in general; for lenders, track applications and funded loans by source, not sessions.

## What decides whether an assistant names a lender?

Platforms say little; studies point to market-specific sources, independent coverage and reputation evidence, which matter more in lending.

**Documented by platforms.** Google says its AI features may use [“query fan-out”](https://developers.google.com/search/docs/appearance/ai-features), issuing related searches across subtopics. A question about consolidating debt with fair credit can pull in separate sources on rates, eligibility and individual lenders. We found no platform documentation on how assistants choose lenders.

**Observed in studies.**

- **The country matters most in lending.** In our country study, the gap between same-country and different-country answers on ChatGPT was 0.254 for high-dependence questions such as insurance, lending and tax, against 0.053 for global products.
- **Review sites settle the trust question.** Asked “is this brand legit?”, assistants in [our reputation study](https://underneath.agency/research/is-it-legit-ai-reputation-study) cited a review or complaint platform in 88.0% of answers; Trustpilot and the BBB together accounted for 61.7% of review-platform citations, and 99.7% of answers flagged at least one problem.
- **Independent mentions track with recommendations.** [Our brand study](https://underneath.agency/research/brand-entity-ai-recommendations-study) found that every tenfold rise in the number of independent sites naming a brand in the cited pages came with 4.7 times the odds of a recommendation, which matters for a lender that only its own site describes.

**Our inference for lenders.** Lending is often licensed state by state and priced by credit profile, so answers lean on sources that explain those details: comparison marketplaces, personal finance publishers and regulators. A lender whose rates, fees, eligibility and licensing are stated plainly gives those sources, and the assistants that read them, accurate material. Complaint patterns, especially about cost surprises, will surface in “is it legit?” answers.

## Which lending rules shape what a platform can say?

The same advertising and fair-lending rules that govern ads; content meant to shape AI answers needs the same compliance review.

We do not offer legal advice, but the rules any lender’s content team works within are public:

- **Truth in Lending advertising.** [Regulation Z](https://www.law.cornell.edu/cfr/text/12/1026.24) requires that an advertised rate of finance charge be stated as an “annual percentage rate,” using that term. Stating certain “triggering terms,” such as the number of payments or the amount of any payment, requires further disclosures.
- **Equal credit opportunity.** [Regulation B](https://www.law.cornell.edu/cfr/text/12/1002.4) bars statements “in advertising or otherwise” that would discourage a reasonable person from applying on a prohibited basis.
- **No approval promises.** Approval always depends on underwriting, and the household survey shows how often applicants are declined or offered less.

Our inference: a page written to be quoted by AI assistants is still marketing. Generative engine optimization (GEO) that inflates approval odds or hides costs is both a regulatory risk and the kind of tactic that can [backfire on a brand](https://underneath.agency/resources/can-geo-backfire-on-your-brand).

## What does GEO look like for a lending platform?

Making the facts borrowers and assistants check accurate, consistent, compliant and confirmed by independent sources.

1. **Rates and costs, plainly.** APR ranges, fees and representative examples on public pages, consistent everywhere, reviewed by compliance.
2. **Eligibility in the open.** Who the product suits and who it does not, including credit, income and state availability.
3. **Answers by need.** Pages for debt consolidation, home improvement, working capital or equipment, each explaining costs and alternatives honestly.
4. **Marketplace accuracy.** Correct, current listings on the comparison sites that assistants cite.
5. **Reputation work.** Answer complaints on Trustpilot and the BBB, especially about costs and repayment terms, and fix the causes.
6. **Licensing and identity facts.** State licenses, the legal entity, partner banks where relevant, and how to verify them, which helps answers to “is it legit?”
7. **Market-by-market checks.** Track answers in each state or country you lend in, because lending answers vary most by market.

Related guides cover [consumer fintech apps](https://underneath.agency/resources/consumer-fintech-apps-customers-ai-search), [fintech startups](https://underneath.agency/resources/fintech-startups-customers-ai-search) and [how to fix wrong brand information in AI answers](https://underneath.agency/resources/fix-wrong-brand-information-in-ai-answers). No one can guarantee that an assistant will recommend a lender, and no visibility work changes who qualifies for a loan.

## Which questions about AI and borrowing remain open?

How often borrowers choose a lender from an AI answer, and whether those borrowers apply and repay differently.

- **No direct measure.** We found no public data on how many borrowers pick a lender because an assistant named it.
- **Vendor surveys.** The consumer AI figures come from Credit Karma, which operates a lending marketplace.
- **Studies outside lending.** Our country study included lending questions, but our reputation and brand studies did not test lenders specifically.
- **Fair-lending questions.** We found no public research on whether AI answers describe loan options differently for different groups of borrowers, a question lenders’ compliance teams may want to watch.

## Where should a lending platform start?

Start by asking assistants your borrowers’ real questions, by product and market, and checking every rate and claim they repeat.

That first check shows which lenders and marketplaces are named for each need, whether your APR ranges, fees and eligibility come back correctly, how “is it legit?” answers describe you, and which sources they rely on. It also shows where answers imply approval odds you would never advertise.

If more qualified applications and funded loans are the goal, [ask us to review how assistants describe your loans](https://underneath.agency/contact). We will test the borrower questions that matter for your products and markets, find inaccurate or noncompliant descriptions, and plan, with your compliance team, the content, marketplace and reputation work that gives assistants accurate evidence. Lenders can read how that work is structured and checked market by market, without any promise about approvals, on our [generative engine optimization service](https://underneath.agency/services/generative-engine-optimization) page.

## Frequently asked questions

### Can GEO get more borrowers approved?

No. Visibility can influence who applies and how well they understand the product; approval depends entirely on the lender’s underwriting and the borrower’s situation.

### Do people really ask ChatGPT where to get a loan?

We found no direct measure. Credit Karma found 66% of generative AI users had used it for financial advice, and 80% of those who acted still validated it first.

### Why do AI answers about loans cite comparison sites?

They explain rates, eligibility and lenders side by side. Our country study also found lending answers vary by market, which favors sources specific to each market.

### Is content written for AI answers subject to lending advertising rules?

Treat it as marketing. Regulation Z’s disclosure rules and Regulation B’s ban on discouragement apply to advertising; your counsel should review any page meant to inform borrowers.

## Sources

- TransUnion (2026-08), [Q2 2026 Credit Industry Insights Report: Consumer Lending](https://media.transunion.com/content/dam/transunion/us/business/collateral/report/1-fs/q2-2026-ciir-consumer-lending-report.pdf)
- Federal Reserve Banks (2026), [Small Business Credit Survey: 2026 Report on Employer Firms](https://www.fedsmallbusiness.org/-/media/project/clevelandfedtenant/fsbsite/reports/2026/2026-report-on-employer-firms/2026-report-on-employer-firms.pdf)
- Board of Governors of the Federal Reserve System (2025-05), [Economic Well-Being of U.S. Households in 2024](https://www.federalreserve.gov/publications/files/2024-report-economic-well-being-us-households-202505.pdf)
- Intuit Credit Karma, via Stacker (2025), [The rise of fin-AI: Why Americans are trusting generative AI with their wallets](https://stacker.com/stories/personal-finance-investing/rise-fin-ai-why-americans-are-trusting-generative-ai-their)
- Legal Information Institute, Cornell Law School, [12 CFR § 1026.24 – Advertising (Regulation Z)](https://www.law.cornell.edu/cfr/text/12/1026.24)
- Legal Information Institute, Cornell Law School, [12 CFR § 1002.4 – General rules (Regulation B)](https://www.law.cornell.edu/cfr/text/12/1002.4)
- Google Search Central (2025), [AI features and your website](https://developers.google.com/search/docs/appearance/ai-features)
- Underneath (2026), [Same question, four countries: do AI recommendations change?](https://underneath.agency/research/ai-recommendations-by-country-study)
- Underneath (2026), [“Is this brand legit?” How AI assistants build a reputation](https://underneath.agency/research/is-it-legit-ai-reputation-study)
- Underneath (2026), [Do Wikipedia and schema make AI assistants recommend a brand?](https://underneath.agency/research/brand-entity-ai-recommendations-study)

---

This is the Markdown twin of https://underneath.agency/resources/lending-platforms-borrowers-ai-search. The HTML page is canonical. Publisher: Underneath, https://underneath.agency/agent. Site index: https://underneath.agency/llms.txt.