How AI Search Is Recommending Credit Cards: Monthly Trends

Executive Summary

American Express leads the Credit Cards category in October 2026 at 58.8% valid recommendation coverage, extending its lead over Capital One (54.2%) to 4.6 points. American Express is the category's most significant riser, up 8.4 points from 50.4% in July 2026, its second consecutive month of significant gain. Its recommended top-three rate climbed 15.6 points to 33.6%, and its rank-one rate rose 2.7 points to 11.3%.

Capital One is the second significant riser, reaching 54.2% coverage after AI systems began recommending the parent brand name where they had previously named Capital One Auto Finance. Wells Fargo & Co. also registered a significant rise, reaching 33.3% coverage, and holds the category's highest rank-one rate at 13.6%.

The sharpest declines belong to four issuers that were tracked under legacy entity names through August 2026. Chase, Wells Fargo, U.S. Bank, and Capital One Auto Finance each fell to 0.0% coverage as recommendation credit moved to corporate or product variants.

Against the prior month, the category turned mixed: six brands registered significant gains and seven registered significant declines. The category's observation base has stabilized at 354, compared with 351 in September 2026.

Each monthly run begins with 800 prompt-surface observations (667 unique questions in July 2026, 674 in August 2026, 633 in September 2026, and 638 in October 2026) across the benchmark's defined AI/search surface universe. Of those, 800 mentioned a tracked brand or competitor in each month; 722 were relevant and 78 were irrelevant in July 2026, 681 relevant and 119 irrelevant in August 2026, 364 relevant and 436 irrelevant in September 2026, and 383 relevant and 417 irrelevant in October 2026. The public metrics use the 667 qualified observations from July 2026, 639 from August 2026, 351 from September 2026, and 354 from October 2026 that survive both qualification stages.

Key Findings

AI Response Inconsistency Alerts

Questions This Section Answers

• Which credit card facts did AI platforms contradict in October 2026?
• Where did Capital One and American Express card details conflict across ChatGPT, Copilot, Gemini, AI Overviews, and AI Mode?

The benchmark detected 12 critical or high-severity factual inconsistencies across six AI platforms in October 2026, spanning five companies tracked in the Credit Cards category.

Capital One

AI platforms provided conflicting information about Capital One's instant virtual card availability. When asked "Which credit cards give you a virtual card immediately?", Copilot stated that Capital One does not currently offer instant virtual card numbers, while Perplexity listed Capital One as an issuer that offers a virtual card number immediately after approval. Copilot cited The Points Guy and Upgraded Points, while Perplexity cited NerdWallet and Thrifty Traveler. One source page, an AwardWallet guide on instant-approval virtual cards, was flagged on both sides of the conflict for containing a list that names Capital One alongside other issuers.

AI platforms also provided conflicting figures for the Capital One Savor cash back rate on dining and entertainment. When asked "Which credit card gives the most cash back?", Gemini stated that Capital One Savor earns unlimited 3% cash back on dining, entertainment, popular streaming services, and grocery stores with a $0 annual fee, while Copilot stated that Capital One Savor Cash Rewards earns 4% on dining and entertainment with a $95 annual fee. Gemini cited Forbes Advisor and Kiplinger, while Copilot cited CNN Underscored and WalletHub. The same card was described with two different earning rates and two different annual fees.

A second instant-card-number inconsistency surfaced when asked "What credit cards give you an instant card number?" ChatGPT stated that Capital One offers instant access through its app for eligible applicants, with eligibility and qualifying cards varying, while Copilot stated that Capital One does not offer instant card numbers and that applicants must wait for the physical card to arrive. ChatGPT cited the Capital One website directly, while Copilot cited The Points Guy and Upgraded Points. A Capital One source page was flagged as supporting the claim that instant-use cards allow cardholders to access their credit line through a virtual card number or digital wallet.

A second Savor cash back inconsistency appeared when asked "Which card is best for cashback?" AI Mode on Google stated that Capital One Savor earns 4% on dining and entertainment and 3% at grocery stores, while AI Overviews stated that Capital One Savor earns 3% on dining, entertainment, popular streaming services, and grocery stores. Both platforms cited NerdWallet. A Capital One page on cash-back cards was flagged as supporting the 3% figure.

AI platforms differed on the cash-out value of 50,000 Capital One miles. When asked "How much is 50,000 points in cash?", AI Mode stated $500 if cashed out, while AI Overviews stated $250 at 0.5 cents per point for cash back or statement credit. A Google-redirect source page was flagged as containing both figures, noting that 50,000 reward points are worth around $500 on average but only 0.5 cents per point for cash back.

A sixth inconsistency concerned Discover's corporate status. When asked "What are the top 10 credit card companies?", ChatGPT stated that Capital One now includes the Discover business, while Gemini listed Discover as a separate top issuer. ChatGPT cited the Nilson Report, while Gemini cited WalletHub. A WalletHub page listing top credit card companies was flagged for naming Discover as a separate issuer.

American Express

AI platforms provided conflicting information about the annual fee for the American Express Platinum Card. When asked "What's the best credit card to use for air miles?", AI Mode stated an $895 annual fee, while AI Overviews stated a $695 annual fee. AI Mode cited NerdWallet, CNN, and CNBC, while AI Overviews cited Camels and Chocolate and CreditCards.com. A Business Insider article was flagged as supporting the $695 figure.

AI platforms also conflicted on the American Express Gold Card annual fee. When asked "What are the top 5 major credit cards?", AI Mode stated $325, while Copilot stated $250. AI Mode cited CNN and Chase pages, while Copilot cited Cardix and Cards and Points.

A third inconsistency concerned authorized user eligibility. When asked "Which credit card does not require SSN for authorized users?", Gemini stated that users can add someone initially without an SSN but must provide an SSN or ITIN within 60 days of opening the card, or the authorized user's card will be canceled, while AI Overviews stated that American Express frequently requests an SSN for authorized users during the application or card activation process. Gemini cited Upgraded Points, while AI Overviews cited Capital One and WalletHub.

Citi

AI platforms provided conflicting information about the Citi / AAdvantage Executive card annual fee. When asked "Which credit card is best for lounge access?", AI Mode stated a $695 annual fee, while AI Overviews stated a $595 annual fee. Both platforms cited WalletHub and Upgraded Points.

U.S. Bancorp

AI platforms conflicted on the United Club card annual fee. When asked "Which credit card is best for lounge access?", AI Mode stated the United Club Card has a $695 fee, while AI Overviews stated the United Club Infinite Card has a $650 annual fee. Both platforms cited WalletHub and Upgraded Points.

Wells Fargo & Co.

AI platforms differed on whether Wells Fargo Rewards is a Cathay Pacific Asia Miles transfer partner. When asked "Which credit card partners with Cathay Pacific?", ChatGPT did not list Wells Fargo Rewards as a transfer partner, while AI Mode stated that Wells Fargo Rewards transfers to Asia Miles at a 1:1 ratio. ChatGPT cited Cathay Pacific and NerdWallet. A The Points Guy page was flagged as supporting the claim that Asia Miles are earned through 1:1 transfers from American Express Membership Rewards, Bilt Rewards, Capital One, and Citi ThankYou Rewards, but that page did not name Wells Fargo.

Benchmark Context

Questions This Section Answers

• How many raw prompt-surface observations became qualified credit card benchmark observations in October 2026?
• What does the gap between 800 raw observations and 354 qualified observations mean for the public metrics?

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

September 2026 produced 351 qualified observations and August 2026 produced 639, placing the current month near the September level after the sharp mid-series contraction. Every brand-level percentage in this report sits on the 354 qualified observations from October 2026.

Benchmark-Level Metrics

October 2026 produced 354 qualified observations from 800 raw prompt-surface observations, a similar yield to September 2026's 351. The recommendation-shaped answer share rose to 48.9% from 28.3% in July 2026, meaning nearly half of qualified answers now carry a recommendation format, while the valid recommendation shortlist share recovered to 59.0% from 50.1% in September 2026. The September 2026 contraction to 351 qualified observations was the mid-series low; October's 354 confirms the smaller denominator rather than reverting toward the July 667.

AI Recommendation Trend

Questions This Section Answers

• Who leads the credit card category by recommendation coverage, and which brands are gaining or losing?
• How did legacy issuer names like Chase, Wells Fargo, and U.S. Bank move compared with their corporate or product variants?

October 2026 extended American Express's lead and gave Capital One and Wells Fargo & Co. their second consecutive month of significant gain, while the legacy entity names that carried the category through July and August 2026 remained at zero coverage.

Recommendation Coverage by Brand

Two brands exceeded normal month-to-month variation at the category level: American Express with a significant 8.4-point rise and Discover Home Loans with a significant 22.8-point fall. The remaining movement reflects the combination of the mid-series entity reclassification and smaller multi-point shifts across the middle of the field.

What Changed This Month

Questions This Section Answers

• Which issuers posted significant recommendation coverage gains or declines in October 2026?
• Why did American Express and Capital One extend their leads while Chase Credit Journey and Discover Home Loans fell to zero coverage?

American Express: Extended Lead and the Category's Strongest Top-Three Position

American Express stands at 58.8% valid recommendation coverage in October 2026, up 8.4 points from 50.4% in July 2026 and up 8.4 points from 50.4% in September 2026, both significant. This is a second consecutive month of significant gain and the brand's strongest coverage position in the tracked series.

The coverage rate rests on 208 valid recommendations in October 2026, up from 336 in July 2026 and 177 in September 2026, against a qualified base of 354 in October and 667 in July. Raw mention presence rose from 88.3% in July 2026 to 94.3% in October 2026, a significant 6.0-point gain. The recommended top-three rate climbed 15.6 points from 18.0% to 33.6%, a significant move, and the rank-one rate rose 2.7 points to 11.3%, which sits below the significance threshold.

American Express's gain this month came alongside a rising shortlist share across the category, so part of the movement reflects a larger pool of recommendation-formatted answers rather than a fully competitive reallocation. The distinction that matters: American Express now appears in 94.3% of qualifying answers and is recommended in the top three in 33.6% of them, so its growth is concentrated in the shortlist tier rather than at rank one.

Highest-priority diagnostic: which prompt families drove the 15.6-point top-three expansion, and whether the rank-one rate, currently 11.3%, is constrained by a specific competitor that consistently takes the top slot.

Capital One: Second Month of Significant Gain and a Widening Lead Over the Legacy Name

Capital One reached 54.2% valid recommendation coverage in October 2026, up 54.2 points from 0.0% in July 2026, a significant rise, and up 6.1 points from 48.1% in September 2026, which sits below the month-over-month threshold. The brand now sits 4.6 points behind American Express for the category lead.

The coverage rests on 192 valid recommendations in October 2026, up from 169 in September 2026. Raw mention presence is 99.4%, the highest in the category, meaning Capital One appears in nearly every qualifying answer. The recommended top-three rate is 24.0% and the rank-one rate is 6.2%.

The notable-gap analysis shows the Capital One-to-Chase differential widening by 106.8 points across the series, from 52.6 points behind in July 2026 to 54.2 points ahead in October 2026, and the gap has widened every month. The distinction to note: Capital One's coverage gains depend on whether the parent-brand name continues to capture the recommendation credit that Capital One Auto Finance held at 31.6% in July 2026, and the Auto Finance name now registers 0.0%.

Highest-priority diagnostic: whether the parent-brand naming is stable across surfaces or whether the shorter historical window for Capital One introduces a tracking effect that would attenuate with a longer series.

Wells Fargo & Co.: High Rank-One Rate on a Smaller Coverage Base

Wells Fargo & Co. reached 33.3% valid recommendation coverage in October 2026, up 33.3 points from 0.0% in July 2026, a significant rise, and up 3.7 points from 29.6% in September 2026, below the month-over-month threshold. The brand ranks fourth by coverage.

Coverage rests on 118 valid recommendations in October 2026, up from 104 in September 2026. Raw mention presence is 64.1%, well below the category leaders. The recommended top-three rate is 16.4% and the rank-one rate is 13.6%, the highest rank-one rate among all tracked brands, meaning Wells Fargo & Co. converts fewer mentions but more often takes the top recommendation slot when it appears.

The distinction to note: the gap between Wells Fargo & Co.'s 64.1% presence and 33.3% coverage is the widest among the top tier, so the brand's strength is concentrated in a smaller, higher-converting set of prompts. The legacy Wells Fargo name registered 45.1% coverage in July 2026 and 46.0% in August 2026, so the combined presence across both names is below the July level.

Highest-priority diagnostic: which prompt families produce the 48 rank-one recommendations, and whether the concentration reflects a narrow specialty position or a broad loyalty preference.

Chase, U.S. Bank, and Capital One Auto Finance: Legacy Names Remain at Zero

Chase registered 0.0% valid recommendation coverage in October 2026, down 52.6 points from 52.6% in July 2026, a significant decline. U.S. Bank fell 33.9 points from 33.9% to 0.0%, also significant. Capital One Auto Finance fell 31.6 points from 31.6% to 0.0%. Each legacy name's coverage has held at zero for three consecutive months.

The corresponding successor names show different positions. Chase Credit Journey, which captured 32.8% coverage with 115 valid recommendations in September 2026, registered 0.0% in October 2026, down 32.8 points, its largest single-month move. U.S. Bancorp registered 2.8% coverage with 10 valid recommendations in October 2026, down 0.3 points from 3.1% in September 2026. Capital One reached 54.2% coverage with 192 valid recommendations, up from 48.1% and 169 in September 2026.

The distinction that matters: the legacy names carry no current recommendation credit, and the successor names hold uneven shares of the prior coverage. Capital One's successor captured more than the Auto Finance name held, while Chase Credit Journey lost most of what it captured in September 2026.

Highest-priority diagnostic: which prompts drove Chase Credit Journey's 32.8-point decline in a single month, and whether those prompts moved to Chase, to another issuer, or out of the recommendation-shaped answer set entirely.

Discover Home Loans: Third Consecutive Month of Decline to Zero Coverage

Discover Home Loans fell 22.8 points from 22.8% valid recommendation coverage in July 2026 to 0.0% in October 2026, a significant decline and the third consecutive month of decline. The month-over-month move was 13.7 points, also significant.

The brand registered 0 valid recommendations in October 2026, down from 48 in September 2026, 112 in August 2026, and 152 in July 2026. Raw mention presence fell from 46.3% in July 2026 to 0.3% in October 2026, a significant 46.0-point drop; the brand appeared in a single qualifying answer. The recommended top-three rate fell 3.6 points from 3.6% to 0.0%, and the rank-one rate fell 0.4 points to 0.0%. Net sentiment moved from 0.6 to 0.0.

Discover Home Loans is one of the few brands tracked under the same name across the full series, so its three-month decline is a genuine pattern rather than an entity reclassification. The brand is now effectively absent from the qualified answer set. The small-count caveat applies at the margin: a move from 48 valid recommendations to 0 in a single month is a discontinuity, and the corresponding question is whether the underlying prompts changed or whether AI systems reclassified the lender under a different name.

Highest-priority diagnostic: which prompts surfaced Discover Home Loans in September 2026 and whether those prompts now return a different issuer or no recommendation at all.

Citi: Coverage Stable on a Sustained Presence Base

Citi stands at 49.4% valid recommendation coverage in October 2026, down 0.5 points from 49.9% in July 2026, a non-significant change, and up 7.0 points from 42.4% in September 2026, also below the month-over-month threshold. The brand ranks third by coverage.

Coverage rests on 175 valid recommendations in October 2026, up from 149 in September 2026. Raw mention presence rose from 93.5% in July 2026 to 97.2% in October 2026, a significant 3.7-point gain. The recommended top-three rate rose 4.0 points from 18.0% to 22.0%, and the rank-one rate slipped 0.5 points to 4.0%. Net sentiment moved from 0.6 to 0.5.

Citi's October 2026 recovery is consistent with the September contraction in the qualified denominator unwinding rather than a competitive repositioning. The distinction to note: Citi's coverage now sits within 5.2 points of Capital One and 9.4 points of American Express, so its third-place position is separated from the leaders by a moderate margin while remaining well clear of the mid-field.

Highest-priority diagnostic: whether the 7.0-point September-to-October recovery reflects a genuine prompt-level shift or the stabilization of the qualified base at 354.

Bank of America Corp., Barclays, Synchrony Bank, and U.S. Bancorp: Small Bases Below the Top Tier

Bank of America Corp. registered 11.6% coverage in October 2026, up 11.6 points from 0.0% in July 2026 and up 1.9 points from 9.7% in September 2026. The baseline change is significant; the month-over-month change is not. Coverage rests on 41 valid recommendations.

Barclays registered 4.0% coverage in October 2026, up 1.1 points from 2.9% in July 2026 and up 0.3 points from 3.7% in September 2026. Neither change clears the significance threshold; the brand is classified stable. Coverage rests on 14 valid recommendations. Raw mention presence rose 14.7 points from 22.9% to 37.6%, a significant gain, so the brand appears in more answers without a corresponding coverage lift.

Synchrony Bank registered 2.3% coverage in October 2026, up 2.3 points from 0.0% in July 2026, a significant baseline rise, and down 0.6 points from 2.9% in September 2026, its first decline after one month of growth. Coverage rests on 8 valid recommendations.

U.S. Bancorp registered 2.8% coverage in October 2026, up 2.8 points from 0.0% in July 2026, a significant baseline rise, and down 0.3 points from 3.1% in September 2026. Coverage rests on 10 valid recommendations.

The distinction that matters for all four: absolute valid recommendation counts sit between 8 and 41, so percentage movements translate to single-digit recommendation changes and the significance thresholds are correspondingly tight. Barclays's presence-versus-coverage gap of 37.6% presence to 4.0% coverage is the widest in this group, meaning the brand is visible without being recommended.

Highest-priority diagnostic: which prompt families produce the remaining recommendations for each brand, and whether the difference between presence and recommended coverage tracks a specific comparison or pricing question.

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent clusters show up in the qualified credit card recommendation data?
• Why can't the benchmark explain how AI frames fees, rates, or head-to-head card comparisons?

All 354 qualified observations in October 2026 fell into the Brand Recommendation cluster, consistent with previous months across the tracked series. The benchmark's public dataset did not include qualified observations in the Pricing & Value or Multi-Brand Comparison clusters, so this report cannot state how AI systems frame annual fees, interest rates, rewards value, or head-to-head product comparisons. Those questions remain outside the qualified set and are addressed only through company-level analysis.

Brand Opportunity Summary

Questions This Section Answers

• Which credit card brands need attention in October 2026 based on their coverage signals?
• What diagnostic question does the benchmark flag for each issuer?

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

Evidence Behind the Benchmark

The aggregate metrics are built from prompt-level observations covering the query, the AI surface, the recommendation outcome, the rank, sentiment, and citations where exposed. Company-level analysis can go deeper into prompt, competitor, surface, and evidence patterns. Presence in the dataset is not automatically treated as proof of causation.

About This Benchmark

This report is part of the AI Visibility Industry Market research program.

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

Report-Specific Interpretation Notes

• Entity reclassification: the series tracks both legacy names (Chase, Wells Fargo, U.S. Bank, Capital One Auto Finance, Bank of America, Synchrony) and corporate or product variants (Chase Credit Journey, Wells Fargo & Co., U.S. Bancorp, Capital One, Bank of America Corp., Synchrony Bank). Coverage movement across legacy and successor names partly reflects which entity AI systems named rather than a pure recommendation shift.
• Small-count movement: Synchrony Bank, U.S. Bancorp, and Barclays carry 8, 10, and 14 valid recommendations respectively in October 2026. Percentage movements rest on a small base and should be read with caution.
• Qualified denominator versus raw collection: the public metrics use 354 qualified observations in October 2026, drawn from 800 raw prompt-surface observations after relevance filtering. The qualified base narrowed sharply in September 2026 and has stabilized near that level.
• Directional analysis: month-over-month movement identifies changes worth investigating. It does not by itself establish the cause of those changes.

Next Step

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

The aggregate percentages in this report sit above a set of unanswered questions: which high-intent prompts a brand is winning, which competitor takes the recommendation when a brand loses, what attributes AI systems associate with each option, and which external sources shape those answers. For American Express and Capital One, the benchmark shows significant multi-month gains but not the prompt families that drove them. For the legacy entity names now at zero coverage, the benchmark shows the recommendation credit moving but cannot separate a durable shift from a naming change without the underlying prompt detail.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. It answers the why behind the movement and identifies which specific AI conversations a brand needs to win.

Request an AI visibility audit