Skip to main content

Citigroup · Financial Services

In AI answers about bank trust and penalties, Citigroup is a roster name, not a lead

When assistants were asked about trustworthy banks, regulatory penalties and banks to avoid, Citigroup appeared in the lists of big names rather than as the subject of a story. The context it carried was mostly compliance and complaints, while its press coverage is about markets and filings.

Measured September 28, 2026ChatGPT, Claude, Gemini, Perplexity and GrokCheck your own brand →

In AI answers

Visibility

32%

Share of sampled answers that named Citigroup. Measured September 28, 2026 · 5 assistants · 25 answers.

Answers naming it

8 / 25

Position by visibility

#4 of 8

In this snapshot only

In the press

News mentions

61,967

PeakMetrics news index, last 30 days.

Citigroup in the headline

2,501

Rank by press volume

#6 of 8

Within this sector

Citigroup gets more AI mentions than press. It ranks #4 of 8 by AI visibility but #6 by news volume. The sampled answers named it more often than its share of coverage would predict.

Named in 3 of 5 questions

Question
Which banks are the most trustworthy?Named by: Claude, Perplexity, Grok
Which financial institutions have the best reputation?Named by: none
Which banks have faced the most regulatory penalties?Named by: ChatGPT, Perplexity, Grok
Which payment networks are the most reliable?Named by: none
Which US bank should customers avoid?Named by: Perplexity, Grok

The evidence

One snapshot, measured September 28, 2026. Live model answers vary between runs, so treat these figures as a single sample of 25 answers, not a ranking or a trend. The questions never named Citigroup.

How the answers frame Citigroup

None of the benchmark questions named a brand. Citigroup surfaced in 8 of 25 answers on 2026-09-28, across three questions: which banks are the most trustworthy, which have faced the most regulatory penalties, and which US bank customers should avoid. Perplexity and Grok each named it in three of their five answers. Claude and ChatGPT named it once each, and Gemini did not name it in this snapshot.

In most of these answers Citigroup sits inside a group of large US banks. Claude listed it among "major national banks" under "generally well-regarded banks." Grok's trust answer placed it last in a Brand Finance-style list that ran from Bank of America through Chase and Wells Fargo. Perplexity's trust answer cited a Lifestory Research study that put Citibank third, behind Chase and Capital One and ahead of Bank of America.

The penalty answers gave the brand its most specific context. ChatGPT named Citigroup with JPMorgan Chase, UBS, Barclays, Royal Bank of Scotland and Bank of America as part of one Fed foreign-exchange enforcement action. Perplexity, citing Good Jobs First's Violation Tracker, listed Citigroup at $27.1B in total penalties and 23 SEC records. Grok gave a different figure for financial offenses, about $12.3B across 236 records. The two answers used different subsets and different numbers, so they should not be read as one total.

In the "avoid" question, Perplexity said Wells Fargo and Bank of America drew the worst customer-reputation signals, and added Citibank as a bank that appears on complaint-based lists. Grok mentioned Citibank among other frequently criticized large banks, after leading with Wells Fargo. Both answers also said no single bank is universally the worst.

The credit tags on these answers point the same way. Citigroup was credited for "Safety & Compliance" in three answers, and three more were tagged "Not Substantive," meaning a name in a list with no distinct claim attached.

…another bank that appears frequently on “worst banks” lists and in complaint-based rankings…
Perplexity, answering “Which US bank should customers avoid?”

What the press is saying

PeakMetrics' news index shows 61,967 mentions of Citigroup in the last 30 days, with 2,501 carrying the brand in the headline. The recent headlines are largely market and product items: a price-target change on Wendy's, warrant reports from Citigroup Global Markets Australia, a short-selling position disclosure from Citigroup Global Markets Limited, and a stock-performance piece asking whether C is outperforming the financial sector.

One headline stands out. It concerns a political story in which the president threatens to sue several companies, including Citigroup, over funding of what the headline calls a "radical left scam." The rest is routine financial-markets coverage, much of it from filings and aggregators.

…all were part of a major Fed foreign-exchange enforcement action totaling more than $1.8 billion.
ChatGPT, answering “Which banks have faced the most regulatory penalties?”

The gap

The press story is about Citigroup as a market participant: analyst actions, disclosures, stock performance. The AI story is about Citigroup as one of several large US banks judged on trust, enforcement history and complaints. Nothing in the sampled headlines resembles the penalty tallies and "worst banks" lists the assistants drew on.

Volume does not translate into visibility either. Among the eight financial-services brands in this snapshot, Morgan Stanley has the largest news count (102,858) but was named in 1 of 25 answers. Bank of America has similar coverage (100,286) and was named in 16. Citigroup sits in the middle on both measures, with 61,967 news mentions and 8 answers.

What to watch

For a Citigroup comms team, the takeaway from this snapshot is about which questions bring the brand up. It appeared when the question was about trust, penalties or avoidance, and the sources behind those answers were ranking surveys, enforcement databases and complaint lists, not the market coverage that fills its news feed.

Two things are worth checking against the cited domains: how the enforcement totals differ between Violation Tracker subsets, and how Citibank is described on complaint-ranking sites. These are one-day results from a small sample, and live answers vary between runs, so they describe this snapshot, not a trend.

The evidence

Where Citigroup appeared, and what it was credited for.

Per-provider split

The same answers split by provider. Each provider contributes a handful of answers, so differences between them are small-sample differences.

  1. ChatGPT
    20%1 / 5
  2. Claude
    20%1 / 5
  3. Gemini
    0%0 / 5
  4. Perplexity
    60%3 / 5
  5. Grok
    60%3 / 5

Appeared in

The benchmark questions whose answers named Citigroup, with the provider that named it. The questions did not mention the brand.

  • Which banks are the most trustworthy?ClaudePerplexityGrok
  • Which banks have faced the most regulatory penalties?ChatGPTPerplexityGrok
  • Which US bank should customers avoid?PerplexityGrok

What the answers credited it for

The topics the answers tied to Citigroup when they named it, as labeled by a classifier run on this site (a Visibility Driver label, not a customer workspace's Smart Category). It describes what the answer discussed, not whether it was favorable.

  • Not Substantive3 answers
  • Safety & Compliance3 answers

Press rank and AI rank

Where Citigroup sits among the 8 financial services brands with news data in this snapshot, on each measure. Ranks only: the two measures use different units and time windows.

Rank by AI visibility

#4 of 8

  1. 4

Share of 25 sampled answers, measured September 28, 2026.

Rank by news mentions

#6 of 8

  1. 6

PeakMetrics news index, last 30 days.

Mention volume is not comparable across sectors, so both ranks are within this sector.

Recent PeakMetrics headlines

A sample of recent coverage from the PeakMetrics news index, covering the last 30 days. Links open the original article.

How this was measured

Measured September 28, 2026. This is a snapshot, not a ranking or a trend. Live model answers vary between runs: in one test, the same brand went from 40% to 30% to 70% visibility across three runs of the same benchmark questions.

This free check asks the same 5 assistants the paid AI Perceptions product benchmarks (ChatGPT, Claude, Gemini, Perplexity and Grok), on the same models, with web search on. Each question was asked once per assistant. It is refreshed periodically, not daily.

It records whether a brand surfaced in an answer and which domains the answer cited. It does not score how favorably a brand was described. The paid platform tracks Favorability daily across all five providers.

The story above was written by an AI model on September 28, 2026 from the stored answers and the PeakMetrics news data for this snapshot. Every quote is checked word for word against a stored answer before it is published.

Your brand

Run the same unbranded questions for your own category.

Check your own brand

Track how AI answers describe your brand every day.

AI Perceptions follows all 5 providers daily, next to the news and social coverage PeakMetrics already monitors.