JPMorgan Chase dominates other banks when it comes to AI banking citations, but regional banks are invisible.
Whenever someone poses banking-related inquiries to AI platforms like ChatGPT, Claude, Gemini or Perplexity, the biggest banks are getting upstaged by third-party comparison hubs and media outlets.
Three websites in particular — Bankrate, Investopedia and Wikipedia — supply 68% of all banking-related AI citations, according to an “AI Visibility” report from communications firm 5WPR. Bank-owned domains, meanwhile, account for less than 7%.
Even within the narrow slice of visibility banks do capture, one name dominates. In response to real consumer questions, such as “best bank near me” or “top bank in [state],” JPMorgan Chase & Co. holds 28.4% of consumer banking AI citation share in the U.S. That’s more than Bank of America (7.1%), Wells Fargo (5.9%), Citi (4.8%), and Capital One (4.2%) combined.
The findings underscore a new reality in the AI era: brand prestige matters less if a chatbot leaves a bank out of the conversation entirely.
JPMorgan Leads Chatbot Citations
“It isn’t accidental,” Ronn Torossian, founder and chairman of 5WPR, told Global Finance in an email. JPMorgan Chase operates more than 4,800 branches across the U.S., but “branch count has almost nothing to do with it,” he added.
Whenever someone consults an AI platform about where to bank, the New York-based firm wins the AI answer outright in only three states: New York, Illinois, and Arizona. Still, Chase owes its AI presence to machine-readable content on its own site, combined with a press footprint that keeps it landing in outlets AI engines already trust, Torossian explained.
But the bank hasn’t locked in the lead just yet. “This is a snapshot, and AI citation patterns shift as engines update retrieval and as competitors invest in the same levers,” he added. “Any bank willing to match that content and structural investment can close the gap.”
Until then, AI assistants will likely continue citing media coverage of banks rather than the banks themselves, he added.
“What these publishers [Bankrate, Investopedia, Wikipedia and also NerdWallet] are doing right is straightforward: comprehensive, frequently updated comparison content, clear schema markup, strong domain authority, and a format built to directly answer the exact questions people and now AI are asking,” Torossian said.
These sites present direct ‘best’ and ‘worst’ rankings of banks, credit cards, and other financial products, giving AI engines structured data to pull from.
Bankrate, for example, feeds answers to specific inquiries about borrowing and connects them with competing lenders. This forces banks to up their game to win over potential customers.
“When banks compete, users get better [interest] rates that help them save money more easily and more effectively,” Bankrate editor-in-chief John Puterbaugh said in an email. “Our commitment to consumer advocacy and helping people get better deals runs across our whole business, and we believe this approach will win even as AI platforms and LLMs continue to evolve.”
What Are Banks Doing Wrong?
Bank websites, by contrast, typically heighten the marketing language to tout their products and offerings. The problem? AI engines ignore that type of content and, instead, identify content that answers specific questions with clarity.
As Andy Mollison, head of search and Innovation at Varn Search Marketing, puts it: AI systems are built around language.
“Vague claims such as ‘we go above and beyond’ provide little useful information,” Mollison said in an email. “A statement such as ‘customers can access support 24 hours a day, seven days a week’ is concrete, verifiable and far more likely to match a user’s query.”
Regulatory and compliance constraints also limit how banks communicate in ways that don’t affect financial-information publishers, Mollison explained. “That often leaves them with less educational content, and more content that is cautious, technical or heavily qualified,” he added.
As a result, publishers have the AI advantage over banks, because they write in language that matches how people actually ask questions.
Regional Banks Face a Discovery Gap
Perhaps the starkest finding from the 5WPR report is this: 22 of the 75 largest U.S. banks registered less than 0.3% citation share. Top bank names — Fifth Third, KeyBank, M&T, Huntington and Regions — barely show up despite their branch networks. Meanwhile, fintech challengers are eating their lunch.
Chime, SoFi, Ally and Discover now out-cite regional banks like PNC, Truist, U.S. Bank and Citizens in AI answers, despite operating with a fraction of the deposit base.
“These five banks registering under 0.3% citation share despite significant size and branch networks isn’t a vanity-metric problem; it’s a discovery problem,” Torossian said. “As more consumers use AI assistants as a first stop for financial research, a bank that’s missing from those answers is missing from consideration at the exact moment decisions are being formed.”
5WPR is careful to frame the index not as a hard count pulled from platform query logs. “Nobody outside those companies has access to that, and any firm claiming otherwise is overselling,” Torossian said. He also acknowledged the report can’t yet tie citation share to account openings or traffic. AI platforms, after all, don’t publish that data. “We’re measuring the front door. We’re not measuring the sale,” he said.
Why AI Invisibility Is Risky
Still, Torossian argued that waiting for proof before investing in AI visibility carries its own risk.
Recall the early days of search engine optimization when Google’s search algorithms transformed how businesses competed online. Companies began investing heavily in web presence during the so-called “SEO Gold Rush” in the early 2000s.
“The brands that showed up first captured the customers,” Torossian said. The ones that waited for proof spent the next decade “trying to catch up,” he added.
“That’s the same bet regional banks made about search fifteen years ago,” he added. “And it’s the same bet that let fintechs out-cite them in AI answers today.”
5WPR isn’t the only agency tracking AI usage among bank consumers. Wells Fargo published a survey in April, alongside the American Bankers Association, reporting that 19% of U.S. adults (and 38% of Gen Z) use AI for financial advice. Two-thirds of those respondents acted on the AI financial suggestions and said those recommendations proved profitable or worthwhile.
In other words, facts matter. Tyler Desjardin, the founder of Pivot Creative Media, a firm that focuses on improving business visibility when it comes to SEO and AI-generated results, advises clients to prioritize just that.
“Brands need to ensure that they provide accurate information so that AI does not have to create something that could be misleading to search engines,” Desjardin said. “Visibility should come from structured information that conveys correct facts, rather than trying to game the system, since AI automatically eliminates any thin or misleading content.”
Anthony Noto covers corporate finance and private credit. Contact him at anoto@gfmag.com