How AI Search Is Recommending Certificates of Deposits: Monthly Trends

Executive Summary

Capital One is the coverage leader in AI-driven certificate of deposit recommendations for October 2026, with valid recommendation coverage of 27.3 percent, ahead of American Express at 24.8 percent, a gap of 2.5 points. That is a narrower margin than the leadership picture a month earlier, when Morgan Stanley held the top position. The category is now a mixed picture: three brands posted significant declines against the July baseline while one posted a significant gain.

The largest prior-month increase belongs to Capital One, whose coverage moved to 27.3 percent in October from zero in September (Up 27.3 points). American Express posted a similar pattern, moving to 24.8 percent from zero in September (Up 24.8 points). Both recoveries follow the brand-name re-identification that reshaped tracking earlier in the series. Morgan Stanley, the September leader, saw the sharpest single-month decline, falling from 32.7 percent in September to 7.4 percent in October (Down 25.3 points, the largest single-month decline of the month, though its move against the July baseline falls within normal month-to-month variation).

Against the July baseline, the category has repositioned. Capital One, American Express, and Bread Savings all declined significantly from their July readings, while Citigroup Inc. is the sole significant riser, up 5.1 points from July to October across a two-month streak. The overall category summary is mixed, with significant movement in both directions and the leader position changing.

Each monthly run begins with 800 prompt-surface observations. In October 2026, those covered 602 unique questions; all 800 mentioned a tracked brand or competitor; 182 were relevant and 618 were irrelevant. The public metrics use the 161 qualified observations that survived both qualification stages. In July 2026, the run also began with 800 prompt-surface observations covering 530 unique questions; 230 were relevant and 570 were irrelevant, and the public metrics used 203 qualified observations.

Key Findings

AI Response Inconsistency Alerts

Questions This Section Answers

• Which banks had the most factual inconsistencies in AI-generated answers this month?
• What types of claims — asset figures, credit card fees, or lounge access — produced the most conflicts?

AI platforms produced 24 critical or high-severity factual inconsistencies across seven platforms in the current measurement period. The conflicts cluster around bank asset figures, credit card annual fees, and lounge access policies, with the largest concentration involving JPMorgan Chase & Co.

JPMorgan Chase & Co.

Seven high-severity conflicts were detected. On the question "What are the top 10 banks?", Google AI Mode stated JPMorgan Chase has $3.46+ trillion in assets while Gemini stated approximately $4.90 trillion, referencing sources including Investopedia and AlphaSense on one side and a Visual Capitalist ranking on the other. On "What are the 25 largest banks in the United States?", ChatGPT stated approximately $4.0 trillion, citing the Federal Reserve and FFIEC, while Gemini stated $4.90 trillion.

On "What are the top 10 banks in the US?", Google AI Mode stated approximately $4.01 trillion, citing eMarketer and U.S. News, while Gemini again stated $4.90 trillion. In a separate pairing on the same question, Gemini stated $4.90 trillion while Copilot stated "$5,015 billion (approximately $5.02 trillion)", citing Wikipedia. Google AI Mode's approximately $4.01 trillion figure and Copilot's approximately $5.02 trillion figure formed another conflicting pair on the same question.

On the stock price question "bank stocks today", Google AI Overviews stated $363.25 (up 0.63%), citing Yahoo Finance, while Google AI Mode stated around $334.47 (up 0.12%), citing Yahoo Finance, Morningstar, and Trading Economics.

On the headquarters question "largest us banks", Google AI Overviews stated JPMorgan Chase is headquartered in Columbus, Ohio, citing the Federal Reserve and Bankrate, while Google AI Mode indicated New York, New York, citing Bankrate. Flagged sources including Experian and DepositAccounts supported the New York listing.

American Express

Six high-severity conflicts were detected across Copilot, Google AI Mode, and Google AI Overviews. On "What is the best travel credit card to have?", Google AI Mode stated the Amex Platinum Card has an $895 annual fee while Google AI Overviews stated $695, both citing NerdWallet and CNBC among their sources. On "What is the best high-end travel credit card?", Google AI Mode stated $895 while Google AI Overviews stated $695, the latter citing a Yahoo Finance article referencing an $895 annual fee.

On "Which is the best credit card to get points for flights?", Google AI Mode stated a $695 annual fee while Google AI Overviews stated $895. On "Which airline credit card is most worth it?", Google AI Mode listed the Delta SkyMiles Gold American Express Card with a $95 annual fee while Google AI Overviews stated the card has a $0 intro fee the first year, then $150.

On "What are the top 5 major credit cards?", Google AI Mode stated the American Express Gold Card has a $325 annual fee while Copilot stated $250. On "Which card gives a free Priority Pass?", Copilot stated the Amex Platinum allows two guests free with $35 per additional guest, while Google AI Overviews stated guest access typically incurs a fee, citing Business Insider and Priority Pass.

Capital One

Four high-severity conflicts were detected across Copilot, Gemini, Google AI Mode, and Google AI Overviews. On "What is the best travel credit card for active duty military?", Google AI Mode stated Capital One generally does not waive annual fees for accounts opened while on active duty, while Google AI Overviews stated the Venture X's $395 annual fee is waived to $0 for active duty, citing Military.com and ChooseFI.

On "How much is 50,000 points in cash?", Google AI Overviews stated $250 (0.5 cents per point) for cash back while Google AI Mode stated a $500 baseline if cashed out, the latter citing American Express and Chase pages. On "How can I get Priority Pass for free?", Google AI Mode stated the Venture X grants unlimited free lounge visits for the cardholder and two guests while Gemini stated complimentary membership is for the primary cardholder only, with additional guests costing $35 each.

On "Which card gives a free Priority Pass?", Copilot stated the Venture X personal card has no free guests and charges $35 per guest, while Google AI Mode stated the card offers unlimited complimentary access for the cardholder and guests, with standard guest fees after initial limits.

Citigroup Inc.

Two high-severity conflicts were detected across Copilot, Google AI Mode, and Google AI Overviews. On "What is the best performing bank stock?", Google AI Mode stated Citigroup is up over 43% on a trailing basis, citing NerdWallet, while Copilot stated a 1-Year Return of +28.7%, citing Sectivia and NerdWallet.

On "top 30 banks in usa", Google AI Overviews stated Citigroup / Citibank has $1,933.6 billion in assets, citing the Federal Reserve and U.S. News, while Google AI Mode stated Citibank has $1.12 trillion in assets, citing Wikipedia and the Federal Reserve.

Goldman Sachs Group

Two high-severity conflicts were detected across ChatGPT, Copilot, Google AI Mode, and Google AI Overviews. On "top 30 banks in usa", Google AI Overviews stated Goldman Sachs Group has $751.7 billion in assets and ranks 5th, citing the Federal Reserve and U.S. News, while Google AI Mode stated $559.81 billion in assets and ranked 6th behind PNC, citing Wikipedia and the Federal Reserve.

On "What are the 5 major U.S. banks?", ChatGPT stated Goldman Sachs is not among the five largest U.S. banks by assets, while Copilot stated Goldman Sachs is one of the five major U.S. banks by total assets, citing Wikipedia and Forbes.

Morgan Stanley

Two high-severity conflicts were detected across Copilot and Gemini. On "What are the 10 commercial banks?", Gemini stated Morgan Stanley is ranked #6 among the 10 largest U.S. commercial banks, citing Wikipedia, while Copilot stated Morgan Stanley is just outside the top 10, citing the Federal Reserve and Wikipedia.

On "What are the top 20 banks in the USA?", Copilot stated $200 billion in assets while Gemini stated $1.58 trillion.

U.S. Bancorp

One high-severity conflict was detected across ChatGPT and Gemini. On "What are the 5 major U.S. banks?", ChatGPT stated U.S. Bancorp is one of the five largest U.S. banks by assets while Gemini stated Goldman Sachs Group is one of the five largest and U.S. Bancorp is not listed, citing Bankrate and the FFIEC.

Benchmark Context

Questions This Section Answers

• How does the qualified analysis set differ from the raw prompt observations collected?
• What does the smaller October denominator mean for how percentage-point movements should be read?

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

The following benchmark-level metrics summarize aggregate recommendation behavior across the qualified set for both the July baseline and the October report period.

Benchmark-Level Metrics

August 2026 carried 204 qualified observations and September 2026 carried 211, before the October count fell to 161. The smaller October denominator means every percentage point represents fewer underlying observations than in the earlier months of the series.

AI Recommendation Trend

Questions This Section Answers

• Which brands posted significant movements against the July baseline?
• How did the leading cluster tighten compared with earlier months?

Leadership returned to Capital One, but the leading cluster is now tighter and the September leader fell back into the pack.

Against the July baseline, only Citigroup Inc. moved enough to register as a significant riser, while Capital One, American Express, and Bread Savings moved enough to register as significant decliners. Morgan Stanley's gain against the baseline falls within normal month-to-month variation, even though its month-over-month drop from September was the largest single-month decline of the period. The remaining brands in the table were classified as stable against the July baseline this month.

What Changed This Month

Questions This Section Answers

• How did Capital One return to the top, and on what kind of base?
• Why did Morgan Stanley's September lead reverse so sharply in October?
• Which brands paired falling coverage with improving placement quality?

Capital One: Back to the top on a narrower base

Capital One's valid recommendation coverage stands at 27.3 percent in October 2026, down 10.6 points from 37.9 percent in July 2026. Against the immediately prior month, coverage moved to 27.3 percent from zero in September (Up 27.3 points), reflecting the resumption of tracking under the brand's common name.

In October, Capital One made 44 valid recommendations out of 161 qualified observations. Its top-three placement rate stands at 13.7 percent, up 4.3 points from 9.4 percent in July, and its rank-one rate is 1.9 percent, down 2.0 points from 3.9 percent in July.

The distinction to notice is that Capital One leads on coverage while its rank-one placement has softened slightly against the baseline. Its raw mention presence rate is 77.6 percent, down 4.7 points from 82.3 percent in July.

Highest-priority diagnostic: Which specific prompt themes and AI surfaces are producing Capital One's coverage lead, and whether the rank-one softening reflects a shift in which CD attributes AI systems associate with the brand.

American Express: A significant recovery month

American Express's coverage stands at 24.8 percent in October 2026, down 12.2 points from 37.0 percent in July 2026, a significant decline against the baseline. Against September, coverage moved to 24.8 percent from zero (Up 24.8 points).

The brand made 40 valid recommendations in October. Its top-three rate is 10.6 percent, up 2.7 points from 7.9 percent in July, and its rank-one rate is 0.6 percent. Raw mention presence is 44.7 percent, down 11.9 points from 56.6 percent in July.

The distinction is that American Express pairs substantial coverage with a declining presence rate. It is being recommended in a meaningful share of qualified observations while being mentioned less often overall than in July.

Highest-priority diagnostic: Which prompt contexts drive American Express's recommendation coverage in the absence of broad mention presence, and which sources surface in those answers.

Bread Savings: Coverage down, placement quality up

Bread Savings's coverage stands at 16.2 percent in October 2026, down 9.4 points from 25.6 percent in July 2026 and down 13.7 points from 29.9 percent in September, both significant movements. Its October figure came from 26 valid recommendations out of 161 qualified observations.

Against the baseline, its top-three rate rose to 7.4 percent from 2.5 percent (Up 4.9 points), and its rank-one rate rose to 5.6 percent from 0.5 percent (Up 5.1 points). Raw mention presence fell to 16.8 percent from 29.6 percent (Down 12.8 points).

The distinction is that Bread Savings is being mentioned less often but placed higher when it is recommended. Top-three and rank-one placement both improved against the baseline even as overall coverage fell.

Highest-priority diagnostic: Which narrower set of prompts still produces Bread Savings recommendations, and whether the higher rank-one rate reflects concentrated strength in a smaller prompt set.

Morgan Stanley: The September leader steps back

Morgan Stanley's coverage stands at 7.4 percent in October 2026, up 4.4 points from 3.0 percent in July 2026, a move that falls within normal month-to-month variation against the baseline. Against September, coverage fell to 7.4 percent from 32.7 percent (Down 25.3 points, significant), the largest single-month move in the series for any brand.

The brand made 12 valid recommendations in October out of 161 qualified observations. Its top-three rate is 3.1 percent, up 3.1 points from zero in July, and its rank-one rate is 1.2 percent. Raw mention presence is 51.5 percent, up 19.5 points from 32.0 percent in July.

The distinction is that Morgan Stanley's presence rate rose against the baseline while its recommendation coverage retreated from September's peak. Visibility and recommendation are moving in different directions for this brand this month.

Highest-priority diagnostic: Whether September's placement levels represented a durable position or a concentration that has since redistributed across Capital One and American Express.

Citigroup Inc.: The category's only significant riser

Citigroup Inc.'s coverage stands at 8.1 percent in October 2026, up 5.1 points from 3.0 percent in July 2026, a significant gain built across a two-month upward streak. The move from September was 2.9 points, which falls within normal month-to-month variation.

The brand made 13 valid recommendations in October. Its top-three rate is 2.5 percent, up 0.5 points from 2.0 percent in July, and its rank-one rate is zero. Raw mention presence is 75.2 percent, up 25.9 points from 49.3 percent in July.

The distinction is that Citigroup is converting a large presence gain into a modest but rising recommendation rate. It appears in three-quarters of qualified observations while its coverage remains below 10 percent.

Highest-priority diagnostic: Why a presence rate of 75.2 percent converts into 8.1 percent recommendation coverage, and which competitor captures the recommendations Citigroup does not.

Goldman Sachs Group: Presence up, coverage down

Goldman Sachs Group's coverage stands at 13.0 percent in October 2026, down 6.2 points from 19.2 percent in July 2026, a move that falls within normal month-to-month variation against the baseline. Against September, coverage rose 0.7 points, extending a two-month upward streak.

The brand made 21 valid recommendations in October. Its raw mention presence rate is 77.6 percent, up 13.1 points from 64.5 percent in July. Its top-three rate is 4.3 percent, up 1.3 points from the baseline.

The distinction is that Goldman Sachs is highly present and modestly recommended. Its presence rate is among the highest in the category, matching Capital One at 77.6 percent and trailing JPMorgan Chase & Co.'s 79.5 percent, while its coverage sits in the middle of the field.

Highest-priority diagnostic: What separates the prompt contexts where Goldman Sachs converts to a recommendation from those where it only appears.

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent cluster captured all qualified CD observations this month?
• What does the benchmark still not answer about how AI compares CD rates and terms across issuers?

All 161 qualified observations in October 2026 fell into the Brand Recommendation cluster. No qualified observations were assigned to the Pricing & Value or Multi-Brand Comparison clusters this month. The public benchmark therefore shows which issuers AI systems recommend, but it cannot yet answer how AI describes rate competitiveness, term value, or how issuers perform when compared directly against one another.

Brand Opportunity Summary

Questions This Section Answers

• Which diagnostic should each brand prioritize to understand its October CD coverage signal?
• Why does the summary flag Citigroup and Morgan Stanley as the most urgent cases to investigate?

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 is not automatically treated as proof of causation.

About This Benchmark

This report is part of the CiteWorks Studio AI Market Discovery 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

• The tracked entity list has alternated between common and formal brand names across the series, with Capital One and American Express appearing under their common names in July and October, and under formal names in August and September. This affects how month-over-month movement is attributed for those two brands.
• Several movements, including U.S. Bancorp at 1.2 percent and Bank of America Corp. at 2.5 percent, come from small absolute counts of valid recommendations. These percentages should be read alongside their underlying counts.
• The qualified denominator fell to 161 observations in October from 211 in September, which changes the weight each observation carries in the percentages.
• Movement between months identifies changes worth investigating; it does not by itself establish what caused 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 hide the questions that matter most for competitive position. Which high-intent CD prompts does each brand win? When Capital One or American Express takes a recommendation, which competitor loses that placement? What product attributes do AI systems associate with each issuer, and which external sources are shaping those answers? None of those answers are visible in month-over-month coverage rates alone.

A company-specific AI visibility audit maps those underlying prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. It turns the benchmark's signal into an actionable diagnosis of where a specific brand wins, where it loses, and what is driving each outcome.

Request an AI visibility audit