Which Publications Do AI Engines Actually Cite? The Ranked List

Three independent 2026 datasets ranked the domains answer engines pull from, and the order looks nothing like a media-prestige list.

Buying media by brand prestige and buying media by citation frequency produce two different target lists. Three datasets published in 2026 measured which domains AI answer engines draw on, and you can read the rankings instead of guessing.

The news outlets AI cites most

LLM Pulse analyzed 5,348,765 citations across 470,380 answers to 16,973 questions between 12 April and 10 July 2026, covering ChatGPT, Google AI Mode, AI Overviews, Gemini, Perplexity and Copilot.

Ranked by raw citation count, the top ten news outlets were TechRadar (14,311), Forbes (12,294), PCMag (11,371), Mashable (11,227), Yahoo (8,396), US News (7,021), CNET (6,353), Tom’s Guide (5,303), TechTarget (3,714) and the New York Times (3,628).

TechRadar alone accounted for 8.21% of all news-outlet citations and Forbes for 7.05%. Below the top ten sat Business Insider (3,468), WIRED (2,427), Fortune (2,157), Consumer Reports (2,090) and CNBC (1,573).

Platforms outrank every publisher

Ahrefs studied over 3 million US queries for its June 2026 report on AI Overviews mention share. YouTube led at 22.9%, Reddit at 18.5%, Facebook at 10.1%, Google’s own properties at 8.8%, Instagram at 5.6%, Quora at 4.7%, English Wikipedia at 4.0%, with TikTok and Amazon tied at 3.3%. Forbes ranked twelfth at 0.8%.

Similarweb found the same shape in roughly 600,000 citation events from January and February 2026. Wikipedia took 13.15% of ChatGPT citations and Reddit 11.97%, with Forbes eighteenth at 1.38%. In Google AI Mode, Fandom led at 7.16%, ahead of Wikipedia at 5.21% and YouTube at 4.91%.

Tech publications take half the news slice

LLM Pulse found that tech publications absorbed 51.9% of all news-outlet citations. That single figure explains why TechRadar, PCMag, Mashable, CNET, Tom’s Guide and TechTarget occupy six of the top nine slots.

If you sell software or hardware, the ranked list doubles as your media list. If you sell legal services, cosmetic procedures or financial advice, the list describes a category you will have trouble entering.

The ranking differs on every engine

Goodie, a vendor that sells answer engine optimization software, examined 31 million AI citations across 11 AI surfaces from October 2025 to July 2026, including 495,000 citations to 37 news domains. News share by engine ran from 23% on Grok down to 5.1% on Claude, with AI Overviews at 19.5%, Gemini at 13.5%, ChatGPT at 11.9%, AI Mode at 10.6% and DeepSeek at 7.9%.

Inside that news share, Goodie recorded Forbes holding 33% of citations, the top five publishers holding 66% and the top ten holding 84%. Concentration that steep means a placement outside the top ten competes for a sixth of the available news citations.

Crawler access reorders the list

The New York Times blocks all 13 major AI user agents. Goodie recorded zero citations to the Times from ChatGPT, zero from Gemini and one from Claude, against 37,642 from Grok, 8,007 from AI Overviews and 4,303 from Perplexity.

The Associated Press blocks every OpenAI crawler and still draws about 80% of its citations from ChatGPT, which Goodie attributes to the AP’s 2023 licensing agreement with OpenAI. A publisher’s robots.txt file and its contract history both move its position in the ranking.

Reading the list as a buyer

Subscribe PR treats these rankings as a filter rather than a target list. A publication’s citation count reflects the questions researchers asked, the engines they sampled and the topics each outlet happens to cover.

Pew Research Center, reviewing 68,879 Google searches from 900 US adults in March 2025, found that 88% of AI summaries cited three or more sources and that the median summary ran 67 words. Multiple sources share a short answer, which leaves room for a mid-tier outlet whose coverage matches the question.

Build the ranking for your own category

You can reproduce this work at a smaller scale. Run thirty of your buyers’ real questions through the engines they use, log every domain each answer cites, then compare your log against the published rankings. The overlap shows you the general-purpose sources. The difference shows you the sources that decide answers in your category.

The free AI visibility audit shows which domains the engines already cite when someone asks about your category.

Subscribe PR is an answer engine optimization and digital PR agency in Phoenix.

SME Paid Under

By Adam

Adam is an owner at Nanohydr8. He really loves comedy and satire, and the written word in general.

Leave a Reply