How AI Search Is Recommending Auto Refinance Loans: Monthly Trends

Executive Summary

Capital One is the new coverage leader in October 2026 at 56.7%, a first-month appearance in the qualified set that opened a 7.8-point gap over second-place LightStream (48.9%). No brand has held the top position across the full series: LightStream led in July 2026 at 45.5%, myAutoloan led in August at 52.6% and September at 46.9%, and Capital One now sits alone at the top. The leadership turnover is the defining feature of the October result.

LightStream posted the largest upward move with a qualifying denominator, rising 6.0 points from 42.9% in September 2026 to 48.9% in October 2026, its sharpest single-month gain in the series. myAutoloan, the September leader, eased 0.7 points to 46.2%, a change inside its normal variation. Both remain in the top three, and the 2.7-point separation between second and third place is the tightest gap in that cluster the category has recorded.

The sharpest decline belongs to a naming transition rather than a competitive loss: Capital One Auto Finance, measured at 37.2% in September, records no qualified coverage in October, while Capital One enters the same vertical at 56.7%. Read as one franchise, the underlying position strengthened; read as separate entities, the public series shows a 41.2-point fall against the July baseline for the Auto Finance label. Gravity Lending and LendingClub remain the clearest competitive decliners, down 8.2 points and 12.1 points respectively from their July baselines.

The category is mixed rather than directional: four brands are significant risers from baseline, three are significant decliners, and four are stable. The notable-gap series widened between Capital One and every major decliner, and the distance between Capital One and LendingClub has widened in every month measured, now standing at 51.4 points.

The July 2026 baseline began with 800 prompt-surface observations (565 unique questions), including 800 brand or competitor mentions, narrowing to 643 relevant prompts and 157 irrelevant prompts for a final qualified benchmark set of 587 observations. The October 2026 report month began with 800 prompt-surface observations (526 unique questions), including 800 brand or competitor mentions, narrowing to 730 relevant prompts and 70 irrelevant prompts for a final qualified benchmark set of 691 observations. August 2026 (420 qualified observations) and September 2026 (623) sit between these two measurement points. All brand-level percentages in this report use the qualified benchmark set as the denominator.

Key Findings

AI Response Inconsistency Alerts

Questions This Section Answers

• Where did AI platforms contradict each other on auto refinance rates and eligibility?
• Which lenders had high-severity pricing or credit-score conflicts in October 2026?

Five critical or high-severity factual inconsistencies were detected across five AI platforms: Copilot, Gemini, AI Mode, AI Mode keywords, and AI Overviews keywords. Every conflict below was flagged at high confidence, and each turned on a figure that a borrower could reasonably act on before applying.

Caribou

Two high-severity pricing conflicts were recorded for Caribou, both on the advertised starting APR. When asked what the best car refinance rate is today, AI Mode stated that Caribou offers starting rates around 4.49% to 4.8% APR, citing Bankrate's auto refinance rates page and Navy Federal Credit Union's auto refinance rates page. Copilot, answering the same question, stated a starting APR of 5.79%, citing the Wall Street Journal's auto refinance rates page, AutoLoanRate's refinance rates page, and LendingTree's rates-by-credit-score page. The flagged source behind the 5.79% figure was AutoLoanRate's refinance rates page, which carries a Caribou refinance marketplace listing at 5.79%.

The second conflict, on the same topic, appeared when the question was framed as the best auto loan refinance rate. Gemini stated Caribou starts as low as 4.18% APR, while Copilot again stated 5.79%. Gemini cited Navy Federal, Bankrate's rates page, and Money.com's auto refinance roundup; Copilot cited Bankrate's refinance rates page, the Wall Street Journal, and AutoLoanRate. The same AutoLoanRate refinance rates page was flagged as the source behind the 5.79% claim.

Auto Approve

One high-severity eligibility conflict was recorded. When asked to compare car refinance rates, AI Mode keywords stated that Auto Approve offers rates as low as 5.29% APR for qualified applicants with 730+ FICO, citing LendingTree, Bankrate, and NerdWallet. AI Overviews keywords, answering the same comparison question, stated that Auto Approve accepts credit scores as low as 480, citing Bankrate and NerdWallet. The two claims describe different borrower profiles for the same lender on the same question, which would produce materially different expectations about who qualifies.

Gravity Lending

One high-severity pricing conflict was recorded. When asked who has the best vehicle refinance rate, Gemini stated Gravity Lending offers rates as low as 3.89%. Copilot, answering the same question, stated that Gravity Lending advertises no fees but gave no comparable rate, implying the brand was not the lowest, and named PenFed as lowest at 4.19%. Copilot cited CNBC Select's auto refinance roundup, AutoLoanRate, and SwitchWize. The flagged source behind the Copilot response was CNBC Select's page describing PenFed's 36-month refinance rates as low as 4.19%, which cannot hold if a 3.89% rate is genuinely available from another lender.

OpenRoad Lending

One high-severity APR range conflict was recorded. When asked whether OpenRoad is good for refinancing, AI Mode stated the brand has starting rates around 5.49%, scaling past 29% for lower-credit borrowers, citing BestCompany, Banks.com, and LendEDU. Copilot stated that OpenRoad's published APRs range from 3.24% to 24.99%, citing RefiMyAuto, the Better Business Bureau profile, and OpenRoad Lending's own pros-and-cons page. Two sources were flagged: a NerdWallet review carrying the 5.49% to 29.90% range behind the AI Mode claim, and a RefiMyAuto review stating OpenRoad's own published range of 3.24% to 24.99% behind the Copilot claim.

Benchmark Context

Questions This Section Answers

• How did the qualified benchmark set change between July and October 2026?
• Why did the qualified observation count rise while the collection frame stayed fixed?

The report separates the raw collection universe from the qualified analysis set. Brand-level recommendation percentages are calculated within the qualified benchmark set only.

Relevance tightened sharply across the series while the collection frame stayed fixed: 157 irrelevant prompts at the July baseline fell to 70 by October, lifting the qualified set from 587 to 691 observations. August 2026, at 420 qualified observations, remains the smallest month in the series, and September 2026 sits at 623. The October qualified set is the largest of the four months.

Benchmark-Level Metrics

The recommendation-shaped answer share crossed the two-thirds mark, moving from 46.8% in July to 50.9% in September to 69.2% in October. AI systems answered a materially larger share of qualified prompts with a recommendation shortlist rather than a factual or comparative response. Over the same span, valid recommendation shortlists rose from 77.2% to 90.7% of qualified observations, meaning responses were not only shaped as recommendations but carried usable shortlists more often.

AI Recommendation Trend

Questions This Section Answers

• Who leads the category in October 2026, and how tight is the top cluster?
• Which brands moved most from the July baseline, and in which direction?
• How should the Capital One and Capital One Auto Finance entries be read together?

A New Leader Emerges While the Top Cluster Tightens to Its Narrowest Spread of the Series

Capital One Auto Finance carried the third-largest baseline position in July and held 37.2% as recently as September, then recorded no qualified coverage in October, while Capital One enters the qualified set at 56.7% in the same month; the two entries are tracked separately in the benchmark, and the transition reads as a naming and entity change rather than a competitive collapse until confirmed otherwise. Below the new leader, LightStream is the only brand whose October move exceeded its month-to-month variation threshold, while four brands are significant risers from baseline and three are significant decliners, so no single mover explains the category's month.

What Changed This Month

Questions This Section Answers

• Which brand changes defined October 2026 compared with the prior month?
• Where did coverage and placement tell different stories for the same brand?

Capital One

Capital One enters the qualified benchmark at 56.7% valid recommendation coverage in October 2026, the current category lead and a 7.8-point margin over LightStream. The brand does not appear in the comparable July baseline, so its October result is a first qualified month rather than a month-over-month gain.

Placement figures are strong alongside coverage: top-three rate stands at 39.4%, raw mention presence at 66.0%, and rank-one rate at 9.6%, with 392 valid recommendations and 272 top-three placements out of 691 qualified observations. The brand reached the highest coverage, presence, and top-three figures in the current month across all tracked brands.

The distinction to notice: this is a broad, high-volume position, not a narrow one. Capital One is surfacing in roughly two of every three qualified prompts and converting most of that presence into shortlist placement, which is a different profile from brands whose coverage rests on a smaller number of top ranks.

Highest-priority diagnostic: Whether the Capital One entry is a continuation of the Capital One Auto Finance position under a shorter brand label, and which of the prior label's prompts carried across.

Capital One Auto Finance

Capital One Auto Finance recorded 41.2% valid recommendation coverage in July 2026, held stable through August at 41.9% and September at 37.2%, then recorded no qualified coverage in October. The baseline-to-current change is down 41.2 points. The prior month-to-current change, from September's 37.2%, is down 37.2 points, and the brand has now declined for two consecutive months by the series classification.

The July baseline placement profile was substantial: raw mention presence 52.6%, top-three rate 22.3%, and rank-one rate 7.0%, with 242 valid recommendations. None of those figures carry into October.

The distinction to notice: coverage reaching zero while the same vertical admits a similarly named brand at the top of the table. The benchmark's own naming change is the most likely explanation, and the public series cannot settle that question on its own. The number to read with care is the 41.2-point fall, which describes a label, not necessarily a franchise.

Highest-priority diagnostic: Whether the prompt set formerly held by Capital One Auto Finance now attributes to Capital One, and which specific prompts and surfaces carried the transition.

LightStream

LightStream rose 6.0 points from 42.9% in September to 48.9% in October, the largest month-over-month gain of any brand with a comparable denominator. Against the July baseline of 45.5%, the brand is up 3.4 points, a change inside its normal variation range.

Placement improved more sharply than coverage. Top-three rate rose from 21.8% in July to 33.0% in October, a significant 11.2-point gain, and rank-one rate rose from 5.3% to 10.4%, a significant 5.1-point gain, with 72 rank-one placements. Raw mention presence eased slightly from 59.5% to 55.1%, a change inside variation. The brand holds 338 valid recommendations.

The distinction to notice: fewer mentions, stronger placement. LightStream is being named somewhat less often while appearing higher in the shortlists where it does appear, which is a quality-of-placement gain rather than a breadth gain.

Highest-priority diagnostic: Which AI surfaces moved LightStream into higher rank positions, and whether that shift is concentrated in a small number of prompts or spread across the question set.

myAutoloan

myAutoloan eased 0.7 points from 46.9% in September to 46.2% in October, a change inside its normal variation range. Against the July baseline of 41.9%, the brand is up 4.3 points, also inside variation.

The September leader ceded the top position but improved on placement strength. Top-three rate rose from 19.1% in July to 28.9% in October, a significant 9.9-point gain, while rank-one rate rose from 4.1% to 6.1%, inside variation, with 42 rank-one placements. Raw mention presence held essentially flat from 48.2% to 48.3%. The brand recorded 319 valid recommendations.

The distinction to notice: coverage held while top-three placement climbed. myAutoloan was already near the top of the field on volume; the October movement is in how high it lands within shortlists rather than how often it appears.

Highest-priority diagnostic: Which prompts moved myAutoloan out of the first coverage position and into the current placement profile, and which surfaces account for the top-three gain.

Caribou

Caribou is the largest riser from the July baseline among brands with a full series, up 7.9 points from 21.8% to 29.7%. The month-over-month move from September's 33.4% was down 3.7 points, inside variation, and the brand has now eased for two consecutive months.

The baseline comparison carries the real signal. Raw mention presence rose 7.9 points from 29.0% to 36.9%, a significant gain. Top-three rate rose 7.7 points from 8.5% to 16.2%, a significant gain. Rank-one rate rose 0.9 points to 2.6%, inside variation. The brand holds 205 valid recommendations and recorded 1 negative mention in the month.

The distinction to notice: a two-month easing at the top of a series that remains well above its baseline. Caribou is not where it was in September on coverage, but it is materially more present and more frequently placed in the top three than it was in July. The 1 negative mention is a single observation and should not be read as a sentiment trend.

Highest-priority diagnostic: Whether the two-month easing reflects prompt rotation or shorter shortlists, and which prompts Caribou stopped appearing in.

Gravity Lending

Gravity Lending is a significant decliner from baseline, down 8.2 points from 36.8% in July to 28.6% in October. The month-over-month change from September's 31.3% was down 2.7 points, inside variation, and the brand is classified as declining for one month after that smaller step.

Raw mention presence fell 8.9 points from 40.0% to 31.1%, a significant drop, and it is the only significant supporting movement. Top-three rate eased 1.8 points from 19.8% to 17.9%, inside variation. Rank-one rate was flat at 4.9%. The brand holds 198 valid recommendations, down from 216 at the July baseline.

The distinction to notice: coverage and presence fell together while placement held. Gravity Lending is being named less often, but when named it is placed about as high as it was in July, which separates a breadth loss from a positioning loss.

Highest-priority diagnostic: Which AI surfaces reduced Gravity Lending mentions, and which brands absorbed those placements.

LendingClub

LendingClub is the largest significant decliner with a full comparable series, down 12.1 points from 17.4% in July to 5.3% in October. The month-over-month change from September's 5.9% was down 0.6 points, inside variation, and the brand has now declined for three consecutive months.

Every supporting metric moved the same direction, and most significantly. Raw mention presence fell 17.8 points from 24.2% to 6.4%, a significant drop. Top-three rate fell 7.8 points from 10.6% to 2.8%, significant. Rank-one rate fell 1.8 points from 2.6% to 0.7%, significant, with 5 rank-one placements. The brand holds 37 valid recommendations, down from 102 at baseline.

The distinction to notice: a three-month decline with no stabilizing metric. Presence, coverage, and placement all fell, and the gap between Capital One and LendingClub has widened in every month measured, now standing at 51.4 points.

Highest-priority diagnostic: Which prompt categories and surfaces stopped surfacing LendingClub, and which competitors received those recommendations instead.

RateGenius

RateGenius is a significant riser from baseline, up 8.6 points from 10.1% in July to 18.7% in October, now extending a three-month upward streak. The month-over-month change from September's 17.0% was up 1.7 points, inside variation.

Placement gains were stronger than coverage gains. Top-three rate rose 6.5 points from 3.2% to 9.7%, a significant gain. Rank-one rate rose 2.4 points from 0.7% to 3.0%, a significant gain, with 21 rank-one placements. Raw mention presence rose 2.7 points from 19.6% to 22.3%, inside variation. Net sentiment improved from 0.6 to 0.8. The brand holds 129 valid recommendations, up from 59 at baseline.

The distinction to notice: the steadiest upward series in the category. RateGenius has gained in each measured month and is the only brand with a three-month streak in either direction that also improved top-three and rank-one placement significantly.

Highest-priority diagnostic: Which surfaces and prompt clusters carried the three-month climb, and whether the gain is durable across the full question set.

RefiJet

RefiJet is a significant riser from baseline, up 4.7 points from 12.1% in July to 16.8% in October. The month-over-month change from September's 18.3% was down 1.5 points, inside variation, ending a single-month hold.

The baseline comparison shows significant movement in presence and placement. Raw mention presence rose 6.7 points from 15.8% to 22.6%, significant. Top-three rate rose 5.4 points from 4.4% to 9.8%, significant. Rank-one rate rose 2.2 points from 0.9% to 3.0%, significant, with 21 rank-one placements. Net sentiment eased from 0.8 to 0.7. The brand recorded 116 valid recommendations and 3 negative mentions.

The distinction to notice: RefiJet now sits just below Auto Approve on coverage while ranking above it on top-three and rank-one rate. Its coverage position and its placement position tell different stories. The 3 negative mentions are a small count and should not be treated as a sentiment shift on their own.

Highest-priority diagnostic: Which prompts account for the gap between RefiJet's coverage rank and its placement rank, and who holds rank-one when RefiJet is named.

Auto Approve

Auto Approve is down 3.9 points from its July baseline of 21.8% to 17.9% in October, a change inside its normal variation range. The month-over-month change from September's 16.9% was up 1.0 point, inside variation, and the brand is now classified as improving for two consecutive months.

Placement held up better than coverage. Top-three rate rose 1.7 points from 9.0% to 10.7%, inside variation, and rank-one rate rose 1.1 points from 1.4% to 2.5%, inside variation, with 17 rank-one placements. Raw mention presence fell 5.3 points from 25.6% to 20.3%, a significant drop, and is the only significant supporting movement. The brand holds 124 valid recommendations.

The distinction to notice: presence fell significantly while coverage and placement held near baseline. Auto Approve is being named in fewer qualified prompts, but the prompts where it is named still place it at roughly the rate seen in July.

Highest-priority diagnostic: Which prompts removed Auto Approve from mention, and whether the lost mentions carried low or high placement value.

OpenRoad Lending

OpenRoad Lending is the smallest tracked brand and stable across the series, up 1.2 points from 5.8% in July to 7.0% in October. The month-over-month change from September's 7.5% was down 0.5 points, inside variation, and the brand has eased for two consecutive months.

Every supporting metric was flat or inside variation. Raw mention presence moved 0.2 points from 7.3% to 7.5%. Top-three rate moved 0.5 points from 2.2% to 2.8%. Rank-one rate was flat at 0.9%, with 6 rank-one placements. The brand holds 48 valid recommendations, up from 34 at baseline.

The distinction to notice: a small base where single placements move the percentage. With 48 valid recommendations, OpenRoad Lending's 7.0% coverage rests on a small count, and individual rank-one placements move the rate more than they would for a larger brand.

Highest-priority diagnostic: Which niche prompts continue to surface OpenRoad Lending, and whether the brand's placement rate holds at a small base.

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent clusters could the October benchmark measure, and which ones could it not?
• What commercial questions remain unanswered because pricing and head-to-head comparison prompts did not qualify?

All 691 qualified observations in October fell into the Brand Recommendation cluster. None qualified for Pricing & Value or Multi-Brand Comparison. This means the public benchmark can currently measure which providers AI systems recommend and how highly they place them, but it cannot yet answer which provider is favored on rate or value, nor which provider wins a direct head-to-head comparison. Those commercial questions require a company-level analysis.

The inconsistency findings above sharpen that gap. AI platforms are already making pricing claims about this vertical, and in five flagged cases they contradicted each other on figures like starting APR and credit-score eligibility. The public benchmark can see that those pricing prompts are in market; it cannot yet attribute them to a cluster that would let a provider track how its own rate claims survive across surfaces.

Brand Opportunity Summary

The benchmark identifies where attention is warranted; a company-level analysis is needed to explain why.

Evidence Behind the Benchmark

The aggregate metrics are built from prompt-level observations: query, surface, recommendation outcome, rank, sentiment, and citations where exposed. Company-level analysis can go deeper into prompt, competitor, surface, and evidence patterns. Source presence in the benchmark is not by itself treated as proof of causation, and a source appearing in an AI answer is not the same as that source endorsing a provider.

Related methodology resources:

• AI Visibility Industry Market
• AI Visibility Industry Market Methodology
• AI Visibility Industry Market Metrics
• AI Visibility Industry Market Standards

Report-Specific Interpretation Notes

• Movement between months identifies changes worth investigating; it does not by itself establish the cause of those changes.
• The qualified benchmark set (691 observations in October) is the denominator for all brand-level percentages; the raw collection frame was larger, and irrelevance fell from 157 prompts at baseline to 70 in October.
• The October qualified set is the largest of the four measured months; August 2026 (420 observations) was the smallest, and comparisons involving August should account for that variation.
• Small counts matter here. LendingClub's 0.7% rank-one rate rests on 5 placements, Auto Approve's 2.5% rests on 17, and OpenRoad Lending's 7.0% coverage rests on 48 valid recommendations; single-digit shifts move these rates more than they would a larger base.
• The Capital One and Capital One Auto Finance entries are tracked separately across the series. The 41.2-point baseline-to-current fall for the Auto Finance label and the 56.7% entry for Capital One should be read together, and the entity relationship is not settled by the public series alone.
• This analysis is directional AI market discovery, not a definitive category ranking.

Next Step

The Public Benchmark Shows Where a Brand Is Winning or Losing. A Company-Level Audit Shows Why.

Beneath the aggregate percentage sit the questions a strategy actually turns on: which high-intent prompts a brand wins, which competitor takes the recommendation when a brand loses, what attributes AI systems associate with each option, and which external sources shape those answers. This month adds a sharper version of that question, because five high-confidence inconsistencies were flagged across five platforms on figures like starting APR and credit-score eligibility. The public benchmark can see the contradiction; it cannot yet show which source or which page produced it.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy, turning the benchmark's signals into an actionable plan.

Request an AI visibility audit