How AI Search Is Recommending VA Loan Lenders

AI-led mortgage discovery is no longer only a “best VA lender” contest. In the May 2026 VA Loans benchmark, AI systems clearly recognize Veterans United Home Loans as a VA-specialist authority, often surfacing it as “best overall” or a top VA-loan option for military borrowers. But when prompts broaden into mortgage lenders, home-loan banks, refinance companies, rates, and digital mortgage experiences, the recommendation field changes.

The benchmark shows a split market: Veterans United owns the strongest VA-specific authority signal, but Rocket Mortgage captures the largest modeled recommendation value across the broader mortgage-lender discovery layer. Rocket Mortgage led the dataset with $283,056 in modeled monthly captured recommendation value, followed by Veterans United at $126,856 and Navy Federal Credit Union at $122,430.

Methodology

1. Market studied: VA loan lenders and adjacent mortgage-lender discovery prompts, including VA-specific, lender-comparison, refinance, rate, and mortgage-pricing questions.
2. Brands/entities included: Veterans United Home Loans, CrossCountry Mortgage, Fairway Independent Mortgage, Freedom Mortgage, loanDepot, Movement Mortgage, Navy Federal Credit Union, New American Funding, Rate, and Rocket Mortgage.
3. Data collection date/window: Report month is May 2026. The uploaded dataset was generated from the Veterans United Home Loans AI Company Index / competitor benchmark packet.
4. AI platforms tested: ChatGPT, Copilot, Gemini, Google AI Mode, Google AI Overviews, and Perplexity.
5. Number of prompts tested: The benchmark includes 1,127 AI search observations. The dataset treats each platform response to a tracked prompt as an observation.
6. Prompt categories: The usable mortgage prompt clusters in the raw observations are Best Mortgage Lenders, Mortgage Lender Comparisons, and Mortgage Pricing and Costs. Buyer stages include consideration, evaluation, and a small number of defensive/pricing-oriented observations.
7. Definition of a mention: A company counted as mentioned when it was present in an AI response, regardless of whether the answer positively recommended it, used it as a comparison anchor, or mentioned it neutrally.
8. Definition of a valid recommendation: A valid recommendation required positive, shortlist-quality treatment in the response. Neutral mentions, comparison anchors, alternatives, and unsupported appearances were not treated as recommendation credit unless the dataset marked them as valid recommendations.
9. Ranking/scoring metrics used: Raw mention presence, valid recommendation coverage, recommended top-three rate, rank-one rate, average recommended rank, positive/neutral/negative visibility, net sentiment by mentions, citation/source patterns, and modeled monthly captured recommendation value. The value metric is a benchmark estimate assigned to eligible top-three recommendations, not revenue.
10. Limitations: This is a point-in-time benchmark. AI answers can change by platform, prompt wording, location, personalization, source updates, and model changes. Modeled captured recommendation value is an estimate for comparative analysis, not revenue, pipeline, or attributable business impact. QA note: the company packet includes stale “Medical Alert Systems” cluster labels in one scope section, so this draft uses the mortgage cluster names from the raw observations and flags the stale labels for cleanup before publication.

Key findings

1. Rocket Mortgage wins the broader AI mortgage-discovery market.
Rocket Mortgage led overall modeled monthly captured recommendation value at $283,056, with a 25.29% top-three recommendation rate and 15.35% rank-one rate. It also had the highest raw mention presence among the tracked brands at 56.52%.

2. Veterans United is the clearest VA-specialist authority.
Veterans United ranked second in modeled value at $126,856, ahead of Navy Federal’s $122,430, even though Navy Federal had broader raw visibility and valid recommendation coverage. Veterans United also showed strong framing quality, with a 0.8686 net sentiment score by mentions and a 10.38% rank-one recommendation rate.

3. Navy Federal is the closest category rival.
Navy Federal Credit Union appeared in 42.15% of observations, earned 26.80% valid recommendation coverage, and captured $122,430 in modeled monthly recommendation value. That puts Navy Federal just behind Veterans United on value and ahead of Veterans United on raw visibility and recommendation coverage.

4. VA-specific prompts are not the whole market.
In VA-labeled prompts, AI systems often grouped Veterans United, Navy Federal, Rocket Mortgage, and PenFed together, with Veterans United frequently framed as “best overall” or a top VA-loan specialist. But broader prompts such as “best mortgage lender,” “best refinance companies,” or “best home loan rates” often shifted recommendation credit toward Rocket Mortgage, loanDepot, Navy Federal, and other generalist lenders.

5. The citation layer is heavily shaped by third-party mortgage sources.
The raw citation records include recurring public sources such as CNBC, NerdWallet, Bankrate, Forbes, mortgage-info.com, and other editorial, review, official, and community-style sources. Those sources appear in AI answers alongside lender recommendations, which makes the public evidence layer a core part of the competitive problem.

What changed in the market

VA loan discovery used to be largely search-led. A military borrower might search Google for “best VA loan lender,” compare a few review pages, visit lender websites, and then submit a lead form.

That journey is becoming more compressed. Borrowers can now ask AI systems to compare VA loan lenders, explain who has low rates, identify the best lender for military borrowers, summarize refinance options, or recommend the best mortgage company before visiting a lender’s site.

That creates a new competitive layer. A lender does not only need to rank in search. It needs to be synthesized correctly inside AI-generated answers. It needs to appear in the right prompt clusters, receive valid recommendation credit, rank high enough to make the buyer shortlist, and be framed with clear reasons to trust it.

For VA loan lenders, this matters because the category is both trust-sensitive and highly specific. Borrowers are not only comparing rates. They are evaluating military-borrower expertise, VA loan specialization, customer experience, online application convenience, eligibility guidance, refinance options, and perceived service quality.

What the benchmark found

The benchmark found a market with two different winners.

Rocket Mortgage is the broad mortgage-discovery winner. It receives substantial AI recommendation credit when prompts broaden from VA-specific language into “best mortgage lender,” “best home loan,” “best refinance company,” “best online mortgage,” and rate-oriented mortgage searches. The AI systems frequently frame Rocket Mortgage around digital experience, speed, online convenience, customer satisfaction, and general mortgage accessibility.

Veterans United is the VA-specialist winner. In direct VA prompts, Veterans United is repeatedly surfaced as a leading VA lender and often framed as “best overall,” “best for VA loans,” or a strong choice for military borrowers. In the dataset, a Google AI Overviews response for “best VA loan lenders” recommended Veterans United, Navy Federal, PenFed, Rocket Mortgage, and New American Funding, with Veterans United at the top of the recommendation list.

Navy Federal is the strongest category-adjacent challenger. Navy Federal benefits from military-borrower relevance, credit-union trust signals, and rate-oriented framing. It does not outrank Veterans United on modeled value in the overall benchmark, but it is close enough to make the category a three-brand race at the top.

loanDepot and CrossCountry Mortgage show value pockets. loanDepot captured $61,221 in modeled monthly recommendation value, while CrossCountry Mortgage captured $25,536. They do not control the VA-specialist story, but they appear in broader mortgage and refinance prompts where AI systems reward general mortgage visibility.

Why visibility is not enough

Veterans United is visible, but visibility alone is not the same as recommendation-stage control.

A lender can be mentioned in an AI answer without being recommended. It can be included as a comparison point but not make the top three. It can be described positively but appear below higher-ranked competitors. Or it can win direct VA prompts while losing broader mortgage prompts where borrowers have not yet used VA-specific language.

That is the central discovery risk in this benchmark. Veterans United is not invisible. It is frequently recognized as a VA-loan specialist. The issue is that the broader AI discovery layer does not always begin with “VA loan lender.” It often begins with questions like:

“What bank is best for a home loan?”
“Who has the best mortgage rates?”
“What is the best company to refinance a mortgage?”
“Which lender is easiest to work with?”
“Who is the best online mortgage lender?”

When prompts shift that way, Rocket Mortgage and Navy Federal capture more of the buyer-shortlist surface area. That is where valid recommendation coverage, top-three placement, and rank-one performance matter more than simple mention presence.

The citation layer

AI systems are not forming these answers from lender websites alone. The citation records show a mix of public evidence sources: editorial lists, review pages, official lender or press sources, finance publishers, community sources, and mortgage comparison content.

In the sampled citation records, AI answers referenced sources such as CNBC, NerdWallet, Bankrate, Forbes, and mortgage-info.com in mortgage-lender recommendation contexts.

That matters because AI systems tend to synthesize from the public evidence layer. If third-party pages consistently describe one lender as “best overall,” another as “best for VA loans,” another as “lowest rates,” and another as “best online experience,” those repeated patterns can shape how AI systems frame the shortlist.

For Veterans United, the citation architecture opportunity is not simply to create more content about VA loans. It is to strengthen the evidence layer around the exact attributes AI systems use when forming recommendations: VA specialization, military borrower experience, rate competitiveness, refinance support, online process quality, trust signals, customer experience, and third-party validation.

What VA loan lenders need to fix

VA loan lenders should not treat AI search as a brand-mention problem. The benchmark shows a more specific set of gaps.

They need to improve valid recommendation coverage, not just raw visibility. They need to earn top-three and rank-one placements in the prompts that create buyer shortlists. They need stronger prompt coverage across both VA-specific and broader mortgage-language searches. They need a more consistent public evidence layer across editorial, review, comparison, official, and community sources. And they need to manage framing quality, so AI systems explain why the lender is the right fit rather than merely listing the brand.

For Veterans United specifically, the strategic question is not “Does AI know Veterans United exists?” It does. The stronger question is: Can Veterans United carry its VA authority signal into broader mortgage, refinance, pricing, and rate-oriented AI discovery moments before Rocket Mortgage or Navy Federal captures the borrower’s shortlist?

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

The VA loan lender market is not only being won by the brand with the clearest category specialization. It is being won by the brands that appear across the full journey of AI-led mortgage discovery.

Veterans United has a strong and defensible VA-specialist position. But the benchmark shows that broader mortgage prompts still route significant recommendation value to Rocket Mortgage and Navy Federal. That creates a practical growth problem: military borrowers may enter the AI discovery journey through general mortgage language before they ever ask a VA-specific question.

For VA loan lenders, the next competitive advantage is not more generic visibility. It is recommendation-stage visibility supported by a stronger citation architecture.

CTA

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

CiteWorks Studio can map where your company appears, where competitors are recommended instead, which prompts carry the most commercial risk, and which public sources are shaping AI-generated recommendations.

Request an AI Visibility Audit or AI Market Discovery Profile.