Parse
IndexPricing
Sign inCheck your brand
  1. Research
Research
Does AI cite sources when it criticizes a brand?

Does AI cite sources when it criticizes a brand?

Published Aug 13, 2026Updated Oct 12, 2026

On this page
  • The finding
  • How we measured
  • Sources
  • More like this
  • The finding
  • How we measured
  • Sources
  • More like this

Dimitry ApollonskyFounder, Parse

DataWhat reviewed means

A criticism can influence how a buyer judges a brand. We wanted to know how often critical statements appear alongside a source attached to that brand.

We examined reviewed statements about ranked brands in ChatGPT Search and Google AI Mode answers from May 24 to July 16, 2026. We compared criticism with brand-level citation attachment; this measures whether a citation is present, not whether it substantiates the criticism.

In 65.9% of the cases where AI criticized a ranked brand, no citation was attached to that brand in the answer. We studied ChatGPT Search and Google AI Mode.

Nearly two in three brand criticisms had no source attached

AI made a specific negative claim in 64,899 ranked brand appearances. In 42,748 of them, or 65.9%, no cited page was attached to that brand in the same answer. The remaining 22,151 had at least one attached cited page.

A criticism audit should start with the uncited cases. An attached page still requires review because attachment does not prove that the page supports the negative claim.

Takeaway

Review uncited criticism first, then verify whether attached pages support the claim.

Criticism was less likely to carry a citation than other brand language

A cited page was attached to 22,151 of 64,899 criticized brand appearances, or 34.1%. Other ranked brand appearances had an attached cited page in 366,725 of 835,166 cases, or 43.9%. Criticism was cited 9.8 percentage points less often.

Overall citation coverage is not a sufficient evidence check. Separate negative claims from other brand language when deciding which answers need source review.

Citation attachment by brand language

  • Specific criticism34.1%22,151 of 64,899
  • No specific criticism43.9%366,725 of 835,166
  • 0%20%40%60%

The engine split was only 2.1 percentage points

ChatGPT Search attached a cited page to 18,261 of 54,056 criticized brand appearances, or 33.8%. Google AI Mode did so for 3,890 of 10,843, or 35.9%. Google AI Mode was higher by 2.1 percentage points.

The uncited-criticism problem appeared on both engines. Audit both, but do not treat the small aggregate difference as evidence that one engine is more accurate.

Citation attachment for brand criticism

  • ChatGPT Search33.8%18,261 of 54,056
  • Google AI Mode35.9%3,890 of 10,843
  • 0%10%20%30%40%

First-place criticism was cited more often than lower-ranked criticism

AI attached a cited page to 3,307 of 6,610 first-place criticisms, or 50.0%. The rate was 36.1% at second or third and 30.3% at fourth or lower. First place exceeded fourth or lower by 19.7 percentage points.

Recommendation rank changes the citation baseline. Compare criticism at similar ranks before deciding that one brand or prompt has an unusual evidence gap.

Citation attachment by recommendation rank

  • First50.0%3,307 of 6,610
  • Second or third36.1%7,454 of 20,676
  • Fourth or lower30.3%11,390 of 37,613
  • 0%20%40%60%

Takeaway

Use rank-specific citation baselines when comparing criticism across brands.

Changing the language definition did not remove the gap

Citation attachment was 34.1% for specific criticism, 26.8% for negative overall tone, 35.4% for reluctant recommendations, and 30.0% for explicit rejection. Each definition left most affected brand appearances without an attached cited page.

A general sentiment score cannot replace claim-level review. Track the exact type of negative language because criticism, reluctance, and rejection describe different decisions.

Citation attachment across language definitions

Reluctant recommendation46,40116,42035.4%
Specific criticism64,89922,15134.1%
Explicit rejection4,2391,27230.0%
Negative overall tone9,5742,56726.8%

Only 3.7% of criticisms had four or more attached sources

Among 64,899 criticized brand appearances, 42,748 had no attached cited page, 12,431 had one, 7,341 had two or three, and 2,379 had four or more. Four or more sources accounted for 3.7% of the total.

Source count helps prioritize review, but it is not a support score. One relevant page can be stronger evidence than several pages that do not contain the claim.

Attached cited pages per brand criticism

  • No attached source65.9%42,748
  • One attached source19.2%12,431
  • Two or three sources11.3%7,341
  • Four or more sources3.7%2,379
  • 0%20%40%60%80%

Industry citation rates differed by 15.9 percentage points

Among displayed industries with at least 500 criticized brand appearances, Sales and Marketing attached a cited page in 398 of 1,033 cases, or 38.5%. Blockchain and Cryptocurrency did so in 134 of 592, or 22.6%. The difference was 15.9 percentage points.

Use an industry baseline before calling a brand's citation rate unusual. The comparison identifies where review may be useful and does not show that industry caused the difference.

Selected industry citation rates

Citation rate · 7 results

  • Sales and Marketing38.5%
  • Education37.1%
  • Artificial Intelligence33.5%
  • Financial Services32.4%
  • Software27.0%
  • Hardware27.0%
  • Blockchain and Cryptocurrency22.6%
  • 0%10%20%30%40%
Explore the data (7 rows)
Sales and Marketing1,03339838.5%
Education88732937.1%
Artificial Intelligence2,38880133.5%
Financial Services4,3871,42232.4%
Software4,9061,32527.0%
Hardware1,09729627.0%
Blockchain and Cryptocurrency59213422.6%

Atlassian and Datadog had the most uncited criticisms

Atlassian had 635 uncited criticisms among 1,000 criticized brand appearances. Datadog had 548 among 915. Alphabet, Notion, ClickUp, Grafana, Elastic, Microsoft, Asana, and Cisco completed the top ten by uncited-criticism count.

This is a review-volume leaderboard, not a brand-quality ranking. Compare each brand's prompt mix, recommendation rank, and cited pages before drawing a reputation conclusion.

Brands with the most uncited criticism

Citation rate · 10 results

  • ClickUp42.9%
  • Datadog40.1%
  • Microsoft37.9%
  • Atlassian36.5%
  • Asana32.9%
  • Notion32.1%
  • Alphabet30.8%
  • Grafana25.7%
  • Cisco18.8%
  • Elastic12.9%
  • 0%20%40%60%
Explore the data (10 rows)
Atlassian1,00063536.5%182
Datadog91554840.1%125
Alphabet66846230.8%437
Notion65144232.1%199
ClickUp67138342.9%121
Grafana49837025.7%110
Elastic37933012.9%84
Microsoft51532037.9%358
Asana46531232.9%120
Cisco34127718.8%75

Takeaway

Use the leaderboard to choose review targets, not to rank brand quality.

The 34.1% citation rate held under stricter matching

The main citation rate was 22,151 of 64,899, or 34.1%. A June 1 start returned 22,149 of 64,896, or 34.1%. Exact brand identities returned 22,074 of 64,899, or 34.0%. Requiring higher-confidence source attachment returned the same 34.1% main rate.

The study counted one brand family once per answer and excluded language that was not validated, brands without a ranked position, and 123 brand-source links without an identifiable cited page. It measures citation attachment, not whether a source mentions the brand, supports the criticism, or makes the criticism accurate.

main result
34.1%

22,151 of 64,899

June 1 through July 16
34.1%

22,149 of 64,896

exact brand identities
34.0%

22,074 of 64,899

higher-confidence attachment
34.1%

22,151 of 64,899

What marketers should do

No cited page was attached in 42,748 of 64,899 observed brand criticisms. Citation attachment varied more by recommendation rank and industry than by engine, and only 3.7% of criticisms had four or more attached sources.

Track specific criticism separately from general sentiment. Audit uncited criticism first. For cited criticism, open the attached page and verify whether it contains and supports the claim. Compare brands at similar ranks within the same industry. Repeat the fixed-window method next quarter before calling a difference movement.

  • AI Overview citation quality: what brands should audit

How we measured

We analyzed 940,245 reviewed AI statements across 900,065 ranked brand appearances, 74,659 brands, 224,593 answers, and 16,161 organic prompts on ChatGPT Search and Google AI Mode from May 24 through July 16, 2026.

of observed brand criticisms had no attached cited page
65.9%

42,748 of 64,899

separated criticism from other ranked-brand language
9.8 points

34.1% versus 43.9% cited

of first-place criticisms had an attached cited page
50.0%

3,307 of 6,610

brands received at least one specific criticism
12,467

Across 16,161 organic prompts

Get the data

Dataset CSVThe metrics behind every figure in this report.

Sources

  1. Semrush: Why 62% of AI citations don't lead to brand mentions (opens in a new tab) · accessed August 13, 2026
  2. Seer: LLM ghost citations (opens in a new tab) · accessed August 13, 2026
  3. BrightEdge: When AI goes negative (opens in a new tab) · accessed August 13, 2026
  4. Evaluating verifiability in generative search engines (opens in a new tab) · accessed August 13, 2026

More like this

Does AI praise match the tone of its cited sources?
The page cited alongside AI praise often did not express a positive view of that brand.
How often does AI recommend a brand with reservations?
AI attached a reservation to a ranked brand in about one in ten appearances.
Does AI criticize its top recommendation?
AI occasionally paired its first recommendation with a specific negative claim about the same brand.
Does AI cite every brand it recommends?
Most answers recommending several brands left at least one of those brands without an attached citation.
Does AI cite sources for its top brand recommendation?
Only about half of first-place recommendations had a citation attached to that brand.
Do AI-cited pages mention the brand?
More than a third of reviewed citations pointed to pages that did not mention the brand linked to that citation.
Do ChatGPT and Google criticize the same brands?
When one engine criticized a brand both ranked, the other usually did not share that criticism.
Which brands does AI tell you to avoid?
AI named brands to avoid much more often when explicitly asked. Unprompted warnings were less common.
AI compliance claims: which brands AI calls SOC 2 or HIPAA compliant
ChatGPT Search and Google AI Mode made 3,349 claims that a named brand meets SOC 2, HIPAA, GDPR, ISO 27001, PCI DSS or FedRAMP. Where every cited page could be read, 77.3% of claims cited a page naming the brand and the standard, and a hand check found 53 of 60 such pages state the status. Backing was 90.9% when the brand’s trust page was cited.

See AI’s criticisms and the sources behind them.

Parse

Index

  • Brands
  • Markets
  • Topics
  • How we measure

For agents

  • Docs
  • Connect
  • Workflows
  • llms.txt

Parse

  • Pricing
  • Research
  • Blog
  • Compare

Company

  • About
  • Contact
  • Corrections

© 2026 Parse, Inc. All rights reserved.

Privacy PolicyTerms of Service