How AI Search Is Recommending Money Market Accounts: Monthly Trends

Executive Summary

Ally Bank holds the coverage lead in October 2026 at 71.0% valid recommendation coverage, a significant rise of 8.6 points from 62.4% in July 2026, and it now sits 5.0 points ahead of Capital One at 66.0%. That gap is narrower than it looks: Capital One entered the tracked series only in September 2026, so its baseline is zero and its movement is a measurement transition, not an organic surge from invisibility.

Ally Bank is the strongest sustained riser in the category, with a two-month upward streak and significant gains on both top-three placement (up 16.5 points from 32.5% to 49.0%) and rank-one wins (up 4.2 points from 14.3% to 18.5%). Capital One posted the largest single-month entry figure in the series at 62.9% in September 2026 and moved a further 3.1 points to 66.0% in October 2026, though that month-over-month move sat inside normal variation.

The sharpest decliner against baseline is Vio Bank-(MidFirst Bank), down 13.5 points from 30.2% in July 2026 to 16.7% in October 2026, a three-month downward streak. Discover Home Loans moved down 10.7 points from 11.0% in September 2026 to 0.3% in October 2026, the sharpest single-month drop in the category this cycle, and its October 2026 result rests on just 2 valid recommendations.

This month's configuration sits at the end of a four-month series in which the qualified observation base has narrowed from 729 in July 2026 to 670 in October 2026. Six brands are classified as significant risers against baseline and six as significant decliners, with the decline side heavily shaped by entity retracking rather than organic loss.

Each monthly run begins with 800 prompt-surface observations (590 unique questions in July 2026, 627 in August 2026, 632 in September 2026, and 617 in October 2026) across the benchmark's defined AI/search surface universe. Of those, 800 mentioned a tracked brand or competitor in each month; 796 were relevant and 4 were irrelevant in July 2026, 758 were relevant and 42 were irrelevant in September 2026, and 761 were relevant and 39 were irrelevant in October 2026. The public metrics use the 729 qualified observations in July 2026 and the 670 qualified observations in October 2026 that survive both qualification stages.

Key Findings

AI Response Inconsistency Alerts

Questions This Section Answers

• Which money market account providers had conflicting APY claims across AI platforms?
• Why does the same institution receive different savings APYs on different AI platforms?
• What source-level conflicts explain the Ally Bank and Capital One rate discrepancies?

The benchmark detected 13 critical or high-severity factual inconsistencies across 8 AI platforms this cycle. Each conflict below pairs two AI systems that answered the same buyer question with mutually incompatible claims about the same institution.

Ally Bank

Ally Bank accounted for 7 of the 13 conflicts, spanning ChatGPT, Copilot, Gemini, AI Mode, and AI Overviews. When asked "Who has the best internet banking?", Copilot stated a 4.10% savings APY while AI Mode stated 3.00% APY on savings. Copilot's answer drew on Tradingpedia, Bankrate, and US News; AI Mode cited Bankrate, NerdWallet, and a SoFi page. A flagged source on Copilot's side carried an Ally Bank listing of "4.10% 0.25% $0 9.4/10."

When asked "Which banks have high yield savings accounts?", ChatGPT stated 3.00% APY citing Fortune, Ally's own rates page, and a Motley Fool review, while Copilot stated 3.80% APY citing US News, Bankrate, and NerdWallet. A flagged source on Copilot's side, a Branchspot roundup, listed Ally Bank's Online Savings Account at 3.80%.

When asked "Who has the best no penalty CDs?", Copilot stated approximately 4.00% APY citing The College Investor, BankBonus, and SwitchWize, while AI Overviews stated 3.40% APY citing NerdWallet, CNBC Select, and Raisin. A flagged source on Copilot's side carried a Marcus no-penalty CD listing at 4.00%, a different institution's product.

A fourth conflict on the same question saw AI Mode state 2.80% APY citing NerdWallet, Raisin, and a Marcus explainer page, against AI Overviews at 3.40% APY citing NerdWallet, CNBC Select, and Raisin. A flagged source on AI Mode's side referenced Ally's own no-penalty CD page showing "2.8% correct as of 10/01/26."

A fifth conflict again on no-penalty CDs saw AI Mode state 2.80% APY while Copilot stated approximately 4.00% APY. Three flagged sources were attached here: a Marcus no-penalty CD page at 4.00%, a College Investor page stating "This account currently yields 2.80% APY," and Ally's own no-penalty CD page reading "2.8% correct as of 10/01/26."

When asked "Which is the best bank to open a new savings account?", ChatGPT stated 3.00% APY citing the Wall Street Journal, Ally's rates page, and Bankrate, while Copilot stated approximately 3.60% to 3.80% APY citing US News, Bankrate, and NerdWallet.

A final conflict concerned account terms rather than rate alone. When asked "3 apy savings account," AI Mode (keyword variant) stated Ally Bank Online Savings offers 3.00% APY with no minimum deposit or monthly fees, while AI Overviews (keyword variant) stated Ally Bank Advantage Savings offers 3.10% APY with an $8 monthly maintenance fee and a $100 minimum opening deposit. AI Mode cited the Wall Street Journal, Investopedia, and a Google redirect; AI Overviews cited an Axos APY calculator, Ally's online savings page, and Sallie Mae's savings page. A flagged source on the AI Overviews side showed "Annual Percentage Yield (APY) 3.1% correct as of 10/01/26 ... No minimum opening deposit," which is itself inconsistent with the $100 minimum described in the same response.

Capital One

Two conflicts centred on Capital One. When asked "Which savings account is best?", AI Mode stated 3.60% APY on the 360 Performance Savings account citing Investopedia, Fortune, and NerdWallet, while Copilot stated 3.10% APY citing US News, SmartAsset, and WealthVieu. Flagged sources on Copilot's side showed "Capital One 360 Performance Savings 3.10% $0 $0" and "Capital One 3.10% APY $0 No monthly maintenance or other fees."

When asked "What savings account does Ramit Sethi recommend?", ChatGPT stated Capital One 360 is his specific favorite and closest match to his personal setup, citing a StudyLib copy of I Will Teach You to Be Rich, a Jordan Harbinger interview, and a PDF of the same book. Copilot stated Ally Bank is his top recommendation and personal choice, with Capital One 360 only a historical favorite at roughly 1% APY, citing Ramit Sethi's own site and GOBankingRates. Flagged sources on both sides carry contradictory quotes from the same author's properties: one page reads "Capital One 360 / ING Direct (This is the one I use)" while another reads "1. Ally Bank (My personal choice) ... Ally is my personal recommendation."

Marcus by Goldman Sachs

Two conflicts involved Marcus. When asked "What is the best bank to have savings with?", ChatGPT stated a current APY of 3.50% citing Bankrate, the Wall Street Journal, and Kiplinger, while Copilot stated APY up to 3.90% citing US News, Bankrate, and Daily Vanguard. A flagged source on Copilot's side, Daily Vanguard, listed Marcus as the "best simple high-APY pick" at 3.90%.

When asked "Is Marcus good for high yield savings?", AI Mode stated the base APY is currently 3.40% citing a Marcus bank review and Marcus's own pages, while Copilot stated APY around 3.65% to 4.50% citing Business Insider, GOBankingRates, and Forbes. A flagged source on Copilot's side showed Business Insider stating "This savings account pays 3.65% Annual Percentage Yield (APY)."

CIT Bank

One conflict involved CIT Bank. When asked "What is the best high yield savings account?", AI Mode stated CIT Bank requires a $100 minimum deposit, citing NerdWallet, a YouTube video, and Bankrate, while Gemini stated the CIT Bank Platinum Savings account requires a minimum balance of $5,000 to earn the top tier APY, citing Bankrate, GreenFi, and Raisin. The two eligibility structures are incompatible for the same account.

Discover Home Loans

One conflict involved Discover Home Loans. When asked "Who took over for Discover Home Loans?", Perplexity stated Capital One acquired or is handling the remaining Discover Home Loans assets, that Discover exited mortgage origination in 2015, and that Capital One's involvement has since been described as winding down or purchasing the unit's remains, citing Discover's home loans page, Banking Dive, and HousingWire. AI Overviews stated Capital One bought Discover Financial Services in May 2025 and shut down the Discover Home Loans division in July 2025, citing Capital One's help centre, Banking Dive, and Discover. Three sources were flagged: NerdWallet stating Discover stopped accepting new home equity and mortgage refinance applications in July 2025, Wikipedia stating Discover was acquired by Capital One in May 2025, and HousingWire reporting a June 16 announcement that Discover was closing its mortgage origination business.

Benchmark Context

Questions This Section Answers

• Why did the qualified benchmark denominator shrink from July to October 2026?
• How many qualified observations and AI surface families back the October 2026 figures?

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

The qualified observation base narrowed across the series: 729 in July 2026, 728 in August 2026, 684 in September 2026, and 670 in October 2026. The August 2026 and September 2026 intermediate months are described here in prose only. The share of prompts filtered out as irrelevant rose from 4 in July 2026 to 39 in October 2026, with the September 2026 peak at 42, so the October 2026 movement is measured against a smaller qualified set than the July baseline.

Benchmark-Level Metrics

The intermediate months matter for interpretation here. The recommendation-shaped answer share was 32.0% in August 2026 and 40.4% in September 2026 before reaching 60.7% in October 2026, so the sharp rise accumulated across the series rather than appearing in a single cycle. The category leader shifted from the predecessor CIT Bank entity in July 2026 to Ally Bank by October 2026.

AI Recommendation Trend

Questions This Section Answers

• Which money market account brands now lead AI recommendation coverage?
• Where has the largest movement occurred between July and October 2026?
• How much of the movement reflects entity retracking rather than organic change?

Ally Bank Leads; the Mid-Tier Is Where the Series Has Moved Most

The top of the category is now a two-brand cluster, with Ally Bank at 71.0% and Capital One at 66.0% separated by 5.0 points. Ally Bank, Capital One, CIT Bank, Sallie Mae, Synchrony Bank, and UFB Direct-Parent Company(Axos Financial, Inc.) each exceeded normal month-to-month variation against baseline, as did Capital One Auto Finance, CIT Bank (Acquiring Company: First Citizens BancShares, Inc), Discover Home Loans, Sallie Mae Bank, Synchrony, and Vio Bank-(MidFirst Bank) on the downside. The category-level change reflects the sum of those movements, several of which are entity retracking rather than organic shifts, with the largest baseline-to-current gaps concentrated in the mid-tier rather than at the top.

What Changed This Month

Questions This Section Answers

• How did Ally Bank convert an unchanged presence footprint into a 16.5-point top-three placement gain?
• Why is Capital One shortlisted often but rarely named first?
• Which brands declined significantly this month, and how much rests on small or retracked counts?

Ally Bank: Broad-Based Gains Across Presence and Placement

Ally Bank reached 71.0% valid recommendation coverage in October 2026, up 8.6 points from 62.4% in July 2026. Against the immediate prior month the move was 0.4 points, inside normal variation, so the series story is a two-month climb rather than a single-month jump.

Raw mention presence rose a modest 1.9 points from 72.7% to 74.6%, inside normal variation. Placement moved far more sharply than presence: the top-three recommendation rate climbed 16.5 points from 32.5% to 49.0%, a significant gain, and the rank-one rate rose 4.2 points from 14.3% to 18.5%, also significant. Valid recommendation coverage moved from 455 valid recommendations in July 2026 to 476 in October 2026.

The distinction to notice: Ally Bank is converting an almost unchanged presence footprint into substantially more shortlist placements. The bank is not appearing in many more conversations; it is being placed in top-three positions far more often when it does appear, and it is winning the top slot more frequently too.

Highest-priority diagnostic: Which surfaces and prompt types account for the 16.5-point top-three gain when presence moved only 1.9 points, and which brands lost shortlist positions in those same prompts?

Capital One: Second Place Held, With the Entry Effect Still Visible

Capital One sits at 66.0% valid recommendation coverage in October 2026, up 3.1 points from 62.9% in September 2026 and therefore inside normal variation month over month. The headline baseline-to-current figure of up 66.0 points is an entity tracking effect, not a 66-point organic gain.

Raw mention presence is 71.9% and the top-three recommendation rate is 38.8%, with a rank-one rate of 9.1%. The brand holds an average recommended rank of 3.27. Valid recommendation coverage rests on 442 valid recommendations in October 2026.

The distinction to notice: Capital One now carries the second-highest top-three placement rate in the category while winning the single top slot at roughly half Ally Bank's rate. The brand is consistently shortlisted but less often named first.

Highest-priority diagnostic: Which prompt types place Capital One in the shortlist but not at rank one, and which brand takes the top slot in those prompts?

Marcus by Goldman Sachs: Recovered Into Third, but the Rebound Is New

Marcus by Goldman Sachs reached 63.6% valid recommendation coverage in October 2026, up only 1.2 points from 62.4% in July 2026 and therefore not significant against baseline. The sharper figure is the immediate prior month: up 11.6 points from 52.0% in September 2026, a significant single-month recovery and a 1-month upward streak.

Raw mention presence fell 5.7 points from 72.3% in July 2026 to 66.6%, a significant decline, so the coverage recovery in October 2026 happened while the bank was being mentioned less often overall. Top-three placement moved 0.9 points from 28.9% to 29.8%, inside normal variation, and the rank-one rate slipped 1.6 points from 7.7% to 6.1%, also inside normal variation. Valid recommendation coverage rests on 426 valid recommendations in October 2026.

The distinction to notice: coverage recovered to roughly baseline while presence remains below baseline. The bank is being recommended in a comparable share of qualified observations but is surfaced in fewer conversations, which points to recommendation credit concentrated in a narrower band of prompts rather than broad visibility recovery.

Highest-priority diagnostic: Which surfaces dropped Marcus by Goldman Sachs from their presence set, and which prompts drove the 11.6-point October 2026 coverage recovery?

Vio Bank-(MidFirst Bank): A Three-Month Decline in Presence Before Placement

Vio Bank-(MidFirst Bank) fell to 16.7% valid recommendation coverage in October 2026, down 13.5 points from 30.2% in July 2026, a significant decline and a three-month downward streak. Against the prior month the move was 1.0 point, inside normal variation, so the erosion accumulated progressively rather than arriving in one cycle.

Raw mention presence dropped 15.8 points from 34.2% to 18.4%, a significant decline, and top-three placement fell 6.1 points from 10.4% to 4.3%, also significant. The rank-one rate held flat at 0.8%. Valid recommendation coverage rests on 112 valid recommendations in October 2026, down from 220 in July 2026.

The distinction to notice: Vio Bank's loss is concentrated in presence, and its rank-one rate was already low at baseline. The bank is being surfaced in roughly half as many conversations and only rarely makes the shortlist when it does appear.

Highest-priority diagnostic: Which brands are now appearing in the prompts where Vio Bank was previously surfaced, and is the shift concentrated on specific AI surfaces?

Discover Home Loans: A Single-Month Collapse on a Very Small Base

Discover Home Loans fell to 0.3% valid recommendation coverage in October 2026, down 5.0 points from 5.3% in July 2026 and down 10.7 points from 11.0% in September 2026. The prior-month move is the sharpest single-month drop in the category this cycle and is significant.

The October 2026 result rests on 2 valid recommendations out of 670 qualified observations. Raw mention presence is 0.3% and top-three placement is 0.0%, meaning the brand did not enter a single top-three shortlist this month. The rank-one rate is 0.0%. Average recommended rank is 4.

The distinction to notice: at 2 valid recommendations, the percentage is arithmetically correct but statistically fragile. A single prompt flipping into or out of the tracked set could move this figure by roughly half a percentage point, so the October result should be read as a near-absence rather than a measured decline to a precise level.

Highest-priority diagnostic: Whether the prompts that surfaced Discover Home Loans in September 2026 are still in the qualified set, and whether the entity is still being resolved by AI systems following the Capital One acquisition timeline flagged in the inconsistency data.

UFB Direct-Parent Company(Axos Financial, Inc.): The Steadiest Mid-Tier Climb

UFB Direct-Parent Company(Axos Financial, Inc.) reached 29.8% valid recommendation coverage in October 2026, up 14.7 points from 15.1% in July 2026, a significant rise and a two-month upward streak. Against the prior month the gain was 10.4 points from 19.4% in September 2026, also significant.

Raw mention presence rose 13.8 points from 17.0% to 30.8%, a significant gain. Top-three placement climbed 9.4 points from 6.9% to 16.3%, also significant, and the rank-one rate rose 4.4 points from 3.2% to 7.6%, also significant. Valid recommendation coverage rests on 200 valid recommendations in October 2026.

The distinction to notice: this is the only mid-tier brand in the series moving significantly across presence, top-three placement, and rank-one wins simultaneously. The improvement is broad rather than concentrated in one layer of the recommendation funnel.

Highest-priority diagnostic: Which prompt types and surfaces drove the simultaneous presence and placement gains, and whether the trajectory holds in the next cycle.

Synchrony Bank and the Entity Transition Cluster

Synchrony Bank reached 25.4% valid recommendation coverage in October 2026, up 25.4 points from a zero baseline and up 7.9 points from 17.5% in September 2026, the latter significant and part of a two-month upward streak. CIT Bank reached 50.3%, up 50.3 points from a zero baseline and up 1.8 points from 48.5% in September 2026, inside normal variation. Sallie Mae reached 11.6%, up 11.6 points from a zero baseline and down 1.1 points from 12.7% in September 2026, inside normal variation.

Capital One Auto Finance, CIT Bank (Acquiring Company: First Citizens BancShares, Inc), Sallie Mae Bank, and Synchrony all registered 0.0% coverage in October 2026 after being tracked under those entity names in July 2026 and August 2026.

The distinction to notice: these paired movements reflect an entity tracking change rather than an organic collapse in AI recommendations, and the successor entities have now been tracked for two months. CIT Bank's successor label captures 50.3% against the predecessor's 57.1% July figure, an 6.8-point gap, while Synchrony Bank captures 25.4% against Synchrony's 29.5% July figure and Sallie Mae captures 11.6% against Sallie Mae Bank's 4.0% July figure. The direction of the gap is not consistent across the pairs.

Highest-priority diagnostic: For each pair, which prompt phrasings determine whether AI systems resolve the recommendation to the parent brand name or the more specific business-line name, and whether the successor figures are now stable across two tracked months.

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent clusters produced qualified observations this cycle?
• What can the benchmark not yet tell brands about rate leadership or head-to-head comparisons?

In October 2026, all 670 qualified observations fell into the Brand Recommendation cluster. Comparison and pricing and cost clusters recorded no qualified observations in either July 2026 or October 2026.

The commercial consequence is specific: the public benchmark can describe which brands AI systems surface and recommend in open-ended money market account questions, but it cannot yet describe which brand is positioned as the rate leader, which brand wins a head-to-head comparison, or whether a brand's fee structure is being represented accurately. Brands whose competitive advantage is rate leadership or direct comparison wins will not see that strength reflected in these metrics. A company-level analysis using pricing and comparison prompts is required to assess those dimensions.

Brand Opportunity Summary

Questions This Section Answers

• Which brands show a significant sustained rise, and which show a sustained decline?
• What is the highest-priority diagnostic for each brand's current AI coverage signal?

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

Evidence Behind the Benchmark

Questions This Section Answers

• What data layer supports the recommendation coverage metrics?
• How should the inconsistency data be interpreted?

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.

The inconsistency data adds a second layer of evidence: where two AI systems answered the same buyer question with incompatible claims about the same institution, the conflict is recorded with the platforms involved, the specific claims, the question asked, and the sources cited. Those conflicts describe divergent representation across surfaces; they do not establish which claim is accurate.

About This Benchmark

This report is part of the LLM Authority Index AI Visibility Market Discovery research program.

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

Report-Specific Interpretation Notes

• Small-count movement: Discover Home Loans at 0.3% coverage rests on 2 valid recommendations in October 2026, and Quontic Bank at 8.5% rests on 57. At these counts a single prompt reclassification can move the percentage materially, so both figures warrant confirmation in the next measurement cycle.
• Qualified denominator versus raw collection: The qualified set narrowed from 729 observations in July 2026 to 670 in October 2026 while the raw collection held at 800, because the irrelevant-prompt count rose from 4 to 39 across the series. Movement is measured within the qualified set.
• Entity retracking: Capital One Auto Finance, CIT Bank (Acquiring Company: First Citizens BancShares, Inc), Sallie Mae Bank, and Synchrony were replaced in the series by Capital One, CIT Bank, Sallie Mae, and Synchrony Bank. Movement between paired entities should be read as a measurement transition, not an organic change in AI recommendation behaviour.
• Directional analysis: The benchmark shows where movement occurred, not why it occurred. Month-over-month movement identifies changes worth investigating; it does not by itself establish the cause of those changes.
• Citation frequency is not endorsement: A flagged or cited source appearing in an AI response is a representation signal, not a statement of accuracy.

Next Step

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

The aggregate percentages raise the questions that matter. For Ally Bank, which high-intent prompts convert an unchanged presence footprint into a 49.0% top-three placement rate, and which competitor loses the shortlist slot when Ally Bank gains it? For Vio Bank-(MidFirst Bank), which prompts and surfaces account for a 15.8-point presence decline, and where did that visibility go? For Marcus by Goldman Sachs, what attributes do AI systems associate with the brand in the prompts where presence held while rank-one wins slipped? And for every brand named in the inconsistency data, which external sources are shaping the competing claims AI systems make about it?

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy.

Request an AI visibility audit