How the answers frame Citigroup
This snapshot, measured on 2026-09-29, covers 250 benchmark questions across ChatGPT, Claude, Gemini, Perplexity and Grok, or 1,250 answers in total. None of the questions named a brand. Citigroup was named in 277 of them. That places it fifth among the 15 financial-services peers in the data, behind JPMorgan Chase, Bank of America, Wells Fargo and Capital One. Grok named it in 96 of its 250 answers, well above ChatGPT (54), Gemini (47), Perplexity (41) and Claude (39).
In the sampled answers, Citigroup usually appears as one entry in a list of big banks. Gemini grouped it with JPMorgan Chase, Bank of America and Wells Fargo as a "Big Four" and described three of them as banks that "required government support or restructuring" in 2008. Grok listed Citigroup's $7 billion settlement as smaller than Bank of America's. Perplexity's market-capitalization table placed it sixth.
Where Citigroup gets a specific role, it tends to be practical and international. Claude's fraud answer credited it with single-use card numbers. Claude and Perplexity both recommended Citibank to travelers and digital nomads as a traditional bank with broad international reach. Claude's remote-work answer said Citibank lets employees split time between remote and office work. Grok's mortgage answer listed Citi among the lenders with the best reputations for fair treatment. ChatGPT's and Gemini's fee answers put Citi in the group of banks that charge a foreign transaction fee.
The credit categories are thin. The most common label was "Not Substantive", assigned to 7 credited mentions, meaning the name appeared without being tied to a clear attribute. The other categories (Price & Value, Safety & Compliance, Reliability & Operations, Labour & Workplace, Financial Performance, Customer Experience) have one or two credits each.
Visibility varied by question intent in this run. It was highest for rewards and perks (40%) and leadership (32%). It was lowest for digital banking experience (9%) and reliability (11%).
Citibank takes fraud protection seriously, automatically alerting customers of potential risks through email and SMS, and is one of the few major banks to implement single-use card numbers.
What the press is saying
PeakMetrics' news index recorded 61,106 mentions of Citigroup over the last 30 days, and 2,583 of them had the brand in the headline. That volume is close to JPMorgan Chase (62,944) and Wells Fargo (65,947), though well below Bank of America (100,429), Goldman Sachs (92,123) and Morgan Stanley (102,303).
The recent headlines in the data are mostly market and finance items. They include stock-performance pieces ("Is C Outperforming the Financial Sector?", "laps the stock market") and ownership filings, such as Citigroup crossing a 3% ownership threshold in Solvay and its markets arm ending a substantial holding in Bapcor. One headline concerns a political story about Trump threatening to sue several companies, Citigroup among them.
The Citibank Plus Account allows users to manage money in multiple currencies, with a global banking network of branches and ATMs in over 160 countries.
The gap
The press story is about stock, holdings and a political fight. The AI story is about consumer features, international reach and workplace flexibility, plus history from 2008. Neither set of themes appears in the other.
Press volume and AI visibility also diverge across the peer set. Capital One has roughly a quarter of Citigroup's news volume (15,891 mentions) but was named in more answers (363). Bank of America and JPMorgan Chase have comparable or larger press footprints and were named far more often. This is one snapshot, but it fits the point that heavy coverage does not translate directly into how assistants describe a company.
While these banks faced significant struggles during the 2008 financial crisis and required government support or restructuring, they have remained core pillars of the U.S. financial system.
What to watch
First, the lack of a strong narrative. The assistants mostly used Citigroup as a filler name in lists, and the most common credit label was "Not Substantive". A comms team could check which attributes it wants credited and whether the sources these assistants cite carry them.
Second, the gaps by question type. Digital banking experience and reliability were the weakest areas in this run. Rewards, perks and leadership were the strongest.
Third, the difference by assistant. Grok named Citigroup far more than the others did, and Claude gave it its most specific roles: fraud tools, international accounts and hybrid work. Since live answers vary between runs, treat these as a baseline to re-measure, not a fixed ranking.