How AI Search Is Recommending VA Loan Lenders

Benchmark-Based Industry Analysis | Powered by LLM Authority Index
Published by CiteWorks Studio

Opening summary

VA loan discovery is no longer happening only through search results, lender comparison pages, paid search, or direct brand queries. Borrowers are also asking AI systems which lender is best, who offers the lowest VA mortgage rates, which bank is strongest for military borrowers, and which mortgage company is easiest to work with.

The May 2026 VA Loan Lenders AI Market Discovery Index shows a split market. Veterans United Home Loans has the clearest VA-specific authority signal. AI systems often frame it as “best overall,” “largest VA lender,” or a leading specialist for VA borrowers. But when prompts broaden from VA-specific language into general mortgage, bank, refinance, rate, or “best lender” language, Rocket Mortgage captures more total modeled recommendation value across the broader mortgage-lender discovery market.

The result is a category-level warning: VA authority matters, but it is not enough by itself. AI-generated recommendations reward brands that are visible across both specialist VA prompts and broader mortgage decision prompts.

Key findings

Rocket Mortgage leads total modeled recommendation value. Across 1,127 observations, Rocket Mortgage captured the highest modeled monthly recommendation value at $283,056, ahead of Veterans United Home Loans at $126,856 and Navy Federal Credit Union at $122,430.

Veterans United ranks second by modeled recommendation value, but first in specialist positioning. Veterans United was repeatedly surfaced as a VA-loan specialist and earned the strongest average recommended rank among major tracked lenders, with an average recommendation rank of 1.19 when it received valid recommendation credit.

Navy Federal is the closest category rival. Navy Federal Credit Union nearly matched Veterans United in total modeled recommendation value and outperformed Veterans United on raw mention presence, valid recommendation coverage, and top-three recommendation rate across the full dataset.

The biggest strategic gap is prompt breadth. Veterans United performs strongly when the buyer asks explicitly about VA loans, VA rates, or veteran mortgage options. Rocket Mortgage wins more often when the prompt shifts to broader mortgage-lender discovery, digital lending, bank comparisons, and general home-loan questions.

Mortgage Pricing and Costs is Veterans United’s strongest cluster. In the Mortgage Pricing and Costs cluster, Veterans United captured $94,663 in modeled recommendation value, ahead of Navy Federal at $54,222 and Rocket Mortgage at $44,505.

What changed in the market

Mortgage buyers are not only searching for lenders. They are asking AI systems to reduce the market into shortlists.

That changes the discovery environment for VA loan lenders. A borrower may ask:

“Who is the best lender for a VA loan?”
“What are VA mortgage rates today?”
“Which bank is best for a home loan?”
“Who has the cheapest mortgage rates?”
“What bank is best for first-time home buyers?”

Those prompts do not all produce the same competitive field. A VA-specific prompt tends to reward specialist authority. A broader home-loan prompt often pulls in generalist mortgage brands, national banks, credit unions, rate-focused lenders, and digital-first mortgage platforms.

This is why Veterans United can be highly authoritative in VA-specific prompts while still losing modeled recommendation value to Rocket Mortgage across the broader AI discovery layer. The category is not only being judged on VA expertise. It is also being judged on convenience, rate competitiveness, digital experience, bank familiarity, refinance utility, and general mortgage comparison visibility.

What the benchmark found

The benchmark analyzed 1,127 observations across ChatGPT, Gemini, Copilot, Perplexity, Google AI Mode, and Google AI Overviews. The tracked prompt clusters were Best Mortgage Lenders, Mortgage Lender Comparisons, and Mortgage Pricing and Costs.

At the overall level, Rocket Mortgage led the market with 56.52% raw mention presence, 31.32% valid recommendation coverage, 25.29% recommended top-three rate, and 15.35% rank-one recommendation rate.

Navy Federal Credit Union ranked second on several visibility and recommendation metrics, with 42.15% raw mention presence, 26.80% valid recommendation coverage, 19.79% top-three rate, and 10.91% rank-one rate.

Veterans United Home Loans had lower total presence than both Rocket Mortgage and Navy Federal, with 20.94% raw mention presence and 14.82% valid recommendation coverage. But its quality of recommendation was strong. Veterans United earned a 12.16% top-three rate, a 10.38% rank-one rate, and the strongest average recommendation position among major lenders at 1.19.

That distinction matters. Veterans United was not the most broadly visible lender, but when AI systems recommended it, they often recommended it with high conviction and VA-specific framing.

The clearest cluster-level story is in Mortgage Pricing and Costs. Veterans United led that cluster by modeled recommendation value, capturing $94,663, ahead of Navy Federal and Rocket Mortgage. It also posted a 16.05% rank-one rate in the pricing cluster and an average recommendation rank of 1.05.

In Best Mortgage Lenders, Rocket Mortgage dominated the broader category. It captured $234,860 in modeled recommendation value in that cluster, compared with $63,347 for Navy Federal, $61,199 for loanDepot, and $31,360 for Veterans United.

That is the core market split: Veterans United owns the clearest VA-specialist signal, but Rocket Mortgage owns more of the broad mortgage-lender discovery layer.

Why visibility is not enough

Raw visibility and recommendation-stage visibility are not the same thing.

A lender can appear in an AI answer as a comparison anchor, an example, a source, a rate reference, or a secondary option without receiving valid recommendation credit. The benchmark separates those outcomes. That distinction is especially important in mortgage, where AI systems often mention several lenders but only clearly recommend a smaller shortlist.

Veterans United’s data shows why this distinction matters. The brand appeared in 20.94% of observations, but valid recommendation coverage was 14.82%. Its top-three rate was 12.16%, and its rank-one rate was 10.38%. That means Veterans United’s opportunity is not simply to be mentioned more often. It is to convert more category-relevant appearances into top-three and rank-one recommendation positions.

Rocket Mortgage shows the opposite pattern at market scale. It had the broadest source footprint inside AI answers and the highest total recommendation value. AI systems frequently framed Rocket as easy to use, digitally strong, fast, and broadly applicable. Those attributes travel well across general mortgage prompts, even when the borrower did not ask specifically about VA loans.

Navy Federal sits between the two. It benefits from military and credit-union relevance, and it competes directly with Veterans United in VA-adjacent prompts. It also performs well in broader bank and rate contexts, which makes it the closest category rival.

The citation layer

The citation layer shows why AI discovery is not just a prompt problem. It is a public evidence problem.

Across the dataset, AI systems cited or surfaced a mix of official, editorial, review, community, government, and other sources. The most frequent cited domains included Bankrate, CNBC, Forbes, NerdWallet, Money, Reddit, The Mortgage Reports, VeteransUnited.com, NavyFederal.org, Military.com, and VA.gov.

This does not prove that any single source caused a specific recommendation. Citation frequency is not endorsement. But the pattern does show the kinds of public evidence AI systems appear to synthesize when forming mortgage-lender answers: lender pages, financial media, review sites, consumer education pages, community discussions, and official/government resources.

For VA loan lenders, that citation architecture is especially important because the category is trust-heavy. Borrowers are evaluating eligibility, rates, fees, benefits, refinancing options, service quality, military specialization, and lender credibility. AI systems need consistent public evidence to connect a lender with the right buying moments.

Veterans United has a strong specialist signal, but the broader source layer needs to support that signal across more non-branded and semi-broad prompts. It is not enough to be known as a VA lender. The evidence layer also needs to reinforce why that specialization matters when a borrower asks about rates, refinancing, first-time buying, lender comparisons, closing experience, or overall mortgage quality.

What brands need to fix

VA loan lenders need to stop treating AI visibility as a brand-mention problem. The benchmark shows that the more valuable issue is recommendation quality.

For Veterans United, the opportunity is to extend VA authority into broader mortgage decision language. The brand already wins when the question is clearly about VA loans. The risk is that AI systems route buyers to Rocket Mortgage, Navy Federal, or other generalist competitors when the language shifts to “best home loan,” “best mortgage company,” “lowest rates,” “first-time buyers,” or “refinance.”

For Navy Federal, the opportunity is to defend and expand its credit-union and military-borrower position. It is close enough to Veterans United in modeled value to be a major category threat, and it has broader visibility across bank and credit-union prompts.

For Rocket Mortgage, the benchmark shows broad AI discovery strength. The brand’s challenge is different: it needs to ensure that its general mortgage visibility does not weaken in VA-specific prompts where specialist lenders can own trust and expertise.

For other lenders, the gap is more fundamental. Brands such as loanDepot, CrossCountry Mortgage, New American Funding, Freedom Mortgage, Fairway Independent Mortgage, Rate, and Movement Mortgage appear in parts of the dataset, but most do not consistently convert visibility into modeled recommendation value at the same level as the top three.

The fix is not to “game” AI systems. The fix is to improve the public evidence layer AI systems summarize: comparison-page coverage, review visibility, owned educational content, VA-specific explanations, rate and refinance resources, third-party validation, source consistency, and prompt-cluster coverage.

How CiteWorks Studio helps

1. Map AI recommendation visibility. Track prompts, platforms, company presence, valid recommendations, top-three and rank-one performance, framing, and citation sources.
2. Identify the sources shaping AI answers. Find the editorial, review, forum, government, directory, owned, and search-visible sources that influence brand framing.
3. Build the citation architecture plan. Strengthen the public evidence layer so AI systems have more accurate, consistent, and persuasive source material to synthesize.

Commercial takeaway

Veterans United has the strongest VA-loan authority signal in the benchmark, but AI discovery is broader than VA-specific authority. The market is being shaped by prompts that blend VA loans, mortgage rates, bank comparisons, digital lending, refinance intent, and first-time-buyer questions.

That is why Rocket Mortgage leads total modeled recommendation value and why Navy Federal remains a close category rival. AI systems are not only asking, “Who is the best VA lender?” They are also answering broader buyer questions where generalist lenders can intercept demand before the borrower narrows the search to VA-specific providers.

For VA loan lenders, the new competitive battleground is recommendation-stage visibility: being present, being clearly recommended, earning top-three and rank-one positions, and being framed by the right public evidence at the moment AI systems form the shortlist.

CTA

Want to know how AI systems are recommending your mortgage brand?

CiteWorks Studio helps lenders understand where they appear, where competitors are recommended instead, which prompts carry the most commercial risk, and which sources are shaping AI-generated recommendations.

Request an AI Visibility Audit or Citation Architecture Review to see how your brand performs across high-intent AI discovery prompts.