How AI Search Is Recommending Consumer Banking: Monthly Trends

Executive Summary

Consumer banking enters its third measurement month with a stable, two-tier hierarchy firmly in place. Regions Bank remains the clear category leader at 17.3% valid recommendation coverage in September 2026, down 1.0 point from its August peak of 18.3%, which the benchmark treats as within normal variation. The gap to Flagstar Bank, the next closest tracked bank at 12.6%, stands at 4.7 points, narrower than the 5.5-point gap recorded in August. No tracked bank posted a significant move this month, and the category's overall shape held steady.

The strongest upward mover this month is Santander Bank, which gained 0.8 points from August to September, from 2.4% to 3.2% coverage, though the benchmark classifies this as within normal month-to-month variation. Santander's two-month rise of 3.2 points from its July baseline of 0.0% is significant. Several banks share smaller gains: City National Bank, First Horizon Bank, and East West Bank each moved up 0.8, 0.8, and 0.1 points respectively. Zions Bank is the sharpest decliner, falling 0.6 points from 0.6% in August back to 0.0% in September, erasing its only measured recommendation presence.

Against the prior month, the category was quiet. Seven of ten tracked banks remain classified as significant risers from their July baseline, but no bank exceeded normal variation between August and September. The benchmark records no significant movement this month, a finding consistent with the emergence of a structured hierarchy in the two prior periods. The category has moved from an empty field in July, to a leader-led ranking in August, to a consolidating pattern in September where gains and losses are small relative to the established order.

Each monthly run begins with 800 prompt-surface observations in July (540 unique questions) and August (596 unique questions), and 700 in September (538 unique questions) across the benchmark's defined AI/search surface universe. Of those, 800 mentioned a tracked brand or competitor in July and August, and 700 in September. Relevant prompts numbered 721 in July, 698 in August, and 605 in September; irrelevant prompts were 79, 102, and 95 respectively. The public metrics use the 392 observations in July, 328 in August, and 277 in September that survive both qualification stages.

Key Findings

Benchmark Context

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 set has contracted in each month of the series, from 392 observations in July to 328 in August to 277 in September. Qualified surface breadth also narrowed from six AI surface families in July and August to five in September, with Gemini absent from the qualified set this month.

Benchmark-Level Metrics

The September response mix across 277 qualified observations included 159 factual answers, 69 recommendation shortlists, 28 ranked lists, 13 comparison analyses, 5 pricing analyses, 2 mixed responses, and 1 cautionary answer. The August intermediate month carried 182 factual answers and 87 recommendation shortlists across its 328 qualified observations.

AI Recommendation Trend

A stable two-tier hierarchy holds, with Regions Bank leading and a mid-tier cluster forming

Regions Bank holds a 4.7-point lead over Flagstar Bank, the only other tracked institution above 12.6% coverage. Behind them, a cluster of four banks sits between 5.1% and 3.2%, while East West Bank, Webster Bank, and Zions Bank remain below 1.0%.

No single bank exceeded normal month-to-month variation between August and September. The category-level stability came from a combination of several smaller movements, with gains of less than one point across most tracked banks offsetting the modest declines at Flagstar Bank, Old National Bank, Pinnacle Financial Partners, Regions Bank, and Zions Bank.

What Changed This Month

Santander Bank

Santander Bank moved from 2.4% coverage in August to 3.2% in September, a gain of 0.8 points, within normal variation. Against its July baseline of 0.0%, the bank's two-month rise of 3.2 points is classified as a significant riser. Its raw mention presence reached 16.6% in September, up from 13.4% in August.

The bank earned 9 valid recommendation mentions out of 277 qualified observations in September, up from 8 out of 328 in August. Its top-three rate held at 0.7% (2 mentions), and it recorded zero rank-one placements for the month, down from 2 in August. Its average recommended rank was 4.0.

Santander Bank maintains the third-highest raw mention presence in the category while converting a modest share into recommendation credit, a gap that persists from the prior month. The bank also recorded 3 negative mentions in September, its first negative sentiment readings in the series, against 11 positive mentions.

Highest-priority diagnostic: Which prompts produced Santander Bank's 9 valid recommendations in September, and where did its negative mentions arise relative to its 16.6% presence?

City National Bank

City National Bank moved from 2.1% coverage in August to 2.9% in September, a gain of 0.8 points, within normal variation. Its two-month rise from a July baseline of 0.0% totals 2.9 points, classified as a significant riser. Its raw mention presence reached 17.3% in September, up from 15.2% in August.

The bank earned 8 valid recommendation mentions out of 277 qualified observations, up from 7 out of 328. Its top-three rate eased to 1.1% (3 mentions) from 1.2% in August, while its rank-one count doubled from 1 to 2 mentions. Its average recommended rank moved from 2.5 to 3.0, a less favorable position than in August.

City National Bank now holds the second-highest raw mention presence in the category at 17.3%, yet converts that visibility into only 2.9% coverage, one of the widest presence-to-recommendation gaps in the vertical. The bank's positive sentiment score rose to 0.3 from 0.26 in August, with 15 positive and zero negative mentions in September.

Highest-priority diagnostic: What prevents City National Bank from converting its near-leading presence into recommendation status at a rate closer to its visibility?

First Horizon Bank

First Horizon Bank moved from 4.3% coverage in August to 5.1% in September, a gain of 0.8 points, within normal variation. Its two-month rise from a July baseline of 0.0% totals 5.1 points, classified as significant. Its raw mention presence reached 13.0% in September, up from 12.8%.

The bank earned 14 valid recommendation mentions in September, matching its August count despite a smaller qualified set. Its top-three rate rose to 4.7% (13 mentions) from 3.7%, and its rank-one count doubled from 3 to 6 mentions. Its average recommended rank stands at 1.69, the strongest average rank among all tracked banks with valid recommendations.

First Horizon Bank continues to convert a mid-tier presence into top-tier placements, with 13 of its 14 valid recommendations landing in the top three. Its rank-one rate of 2.2% now trails only Regions Bank in the category.

Highest-priority diagnostic: What distinguishes First Horizon Bank's prompts such that most of its valid recommendations land in the top three, and can that pattern extend beyond its current coverage?

Zions Bank

Zions Bank fell from 0.6% coverage in August to 0.0% in September, a decline of 0.6 points, within normal variation and its streak now shows one month of decline. The bank recorded zero valid recommendation mentions out of 277 qualified observations in September, down from 2 out of 328.

Its raw mention presence rose to 6.9% in September, up from 4.6%, meaning the bank appears more often in AI responses without earning recommendation credit. Zions Bank registered 19 neutral mentions and zero positive or negative mentions in September, compared with 2 positive mentions in August.

The distinction between visibility and recommendation is sharp for Zions Bank: it is present in nearly 7% of qualified AI responses yet holds no valid recommendations. The single rank-one placement it earned in August did not repeat.

Highest-priority diagnostic: Which prompts produced Zions Bank's August rank-one recommendation, and why did its increased presence in September yield zero recommendation credit?

Flagstar Bank

Flagstar Bank eased from 12.8% coverage in August to 12.6% in September, a decline of 0.2 points, within normal variation. Its two-month rise from a July baseline of 0.0% totals 12.6 points, still classified as significant. Raw mention presence rose to 17.7% in September from 16.8%.

The bank earned 35 valid recommendation mentions out of 277 qualified observations, down from 42 out of 328. Its top-three rate eased to 5.8% (16 mentions) from 6.1%, and its rank-one count fell from 6 to 1 mention. Its net sentiment score of 0.80 remains the strongest in the vertical, with 39 positive mentions and zero negative.

Flagstar Bank holds the second position with a wide gap to the third-place cluster but a narrowing one to Regions Bank. The drop in rank-one placements from 6 to 1 is the most notable shift beneath an otherwise stable headline figure.

Highest-priority diagnostic: Which prompts gave Flagstar Bank rank-one placement in August, and why did that pattern dissipate in September?

Buyer-Intent Interpretation

In September 2026, all 277 qualified observations landed in the Brand Recommendation cluster. No qualified observations fell into the Pricing & Value or Multi-Brand Comparison clusters, consistent with the prior month despite the response-type distribution showing 13 comparison analyses and 5 pricing analyses collected. Those responses did not survive qualification for the public benchmark. The public metrics therefore answer which banks AI systems recommend, but they cannot yet answer how those systems frame pricing tradeoffs or head-to-head comparisons between banks in this vertical.

Brand Opportunity Summary

The benchmark identifies where attention is warranted across all ten tracked banks; a company-level analysis is needed to explain why the patterns differ so sharply between presence and recommendation credit.

Evidence Behind the Benchmark

The aggregate metrics are built from prompt-level observations capturing the query, the AI surface that answered it, whether the response included a recommendation, the rank of any recommended brand, the sentiment of the response, and citations where exposed. Company-level analysis can go deeper into prompt, competitor, surface, and evidence patterns. Presence in the benchmark is not automatically treated as proof of causation.

About This Benchmark

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

• AI Industry Market Discovery Methodology
• AI Industry Market Discovery Metrics
• AI Industry Market Discovery Standards

Report-Specific Interpretation Notes

• Movement from July to September is measured against the qualified benchmark set (392 observations in July; 277 in September), not the raw collection universe of 800 and 700 prompts respectively.
• The August intermediate month (328 qualified observations) is referenced in prose where it clarifies the September result; comparison tables use July and September only.
• Small counts matter in this vertical. Regions Bank's 48 valid recommendations and Flagstar Bank's 35 are meaningful; Webster Bank's 1, East West Bank's 2, and Zions Bank's 0 are included because they represent the full picture for those brands.
• The July baseline of 0.0% coverage across all banks reflects a category with no recommendation-shaped AI responses that month, not a failure of the tracked brands.
• The benchmark records what AI systems surfaced in each month; it does not explain why those systems produced those outputs.
• Attribution: the LLM Authority Index is the benchmark and research authority; CiteWorks Studio provides interpretation, strategy, and remediation as a separate function. No movement reported here is attributed to CiteWorks activity.
• Recommendation coverage measures presence in a recommendation shortlist, not endorsement. A bank can appear in a recommendation without being the stated best option.

Next Step

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

The aggregate percentages raise questions the benchmark alone cannot answer. Which high-intent prompts is Regions Bank winning, and which competitor takes the recommendation when it loses? Why did Flagstar Bank's rank-one placements collapse from 6 to 1 while its presence held steady? What attributes do AI systems associate with First Horizon Bank that drive its strongest-in-category average rank? Why does City National Bank appear in 17.3% of AI responses yet earn recommendation credit in only 2.9%? Which external sources shaped the answers that removed Zions Bank from recommendation shortlists entirely? These are the questions beneath the surface of the trend line.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. The benchmark shows where the market's attention is moving; an audit explains the mechanics behind that movement and where a single brand can act.

Request an AI visibility audit