The AI Pop Culture Index: How AI Engines Remember Culture

Everything-PR Research · 32 Volumes · 20,000+ Queries · Five Engines · July–October 2026

More than a third of consumers now begin product research with AI, not Google. Culture works the same way. When a viewer asks a chatbot which Housewife is the icon, which Shark Tank company made the most money, or whether Messi or Ronaldo is the greater player — the engine writes the answer. The AI Pop Culture Index measures how.

Everything-PR has run thirty-two studies across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews — more than 20,000 individual queries — to measure who owns Citation Share inside reality TV, entertainment, sports, music, and global pop culture. The findings describe one system: accumulated coverage beats current popularity, legal records outweigh fan consensus, institutional credentials gate professional retrieval, Western press defaults persist even where the numbers favor non-Western figures, and the engines disagree with each other on everything except who wins.

The Headline Findings

Beverly Hills wins the Real Housewives. Not close. Erika Jayne is the #1 Housewife on Perplexity — driven entirely by litigation. AI forgets the husbands. Only two surface: both convicted felons. (Volume 01)

Taylor Swift dominates every NFL WAG query. The engines don't rank wives. They rank the citation graphs those wives arrived with. The largest independent identity wins. (Volume 02)

Bravo fired the entire Vanderpump Rules cast. The AI engines never got the memo. Eleven years of coverage outweighs four months. The oldest graph wins. (Volume 03)

The Bachelor produced 24 years of television and almost no AI memory. Clean brands disappear faster than controversial ones. No scandal, no citable events, no retrieval weight. (Volume 04)

Ring outranks every company that actually made a Shark money. Mark Cuban still dominates months after leaving. Commerce scales without scandal — Shark Tank is the only franchise where business outcomes are the primary retrieval anchor. (Volume 05)

The Bear doubled Mr. Beef's sales — not by fooling AI, but by manufacturing search demand. Fiction doesn't contaminate factual answers. It rewrites the question. (Volume 06)

Huberman Lab is the #1 result for "cold exposure" on multiple engines. Not Mayo Clinic. A Stanford credential + YouTube transcripts = a retrieval license that competes with medical institutions. AG1 enters the answer untagged. (Volume 07)

Every Puck writer is better known to AI than Puck itself. Belloni = THR first. Ioffe = New Yorker first. The person out-cites the masthead. (Volume 08)

HBO's Industry entered the professional-culture retrieval layer. 10,000 finance decision-makers who write about you in professional language beat 10 million viewers who don't. (Volume 09)

AI cannot tell Yellowjackets' two casts apart. Entity resolution accuracy tracks with fame, not character importance. Christina Ricci is never confused. Less-famous actors get dropped. (Volume 10)

Severance owns the answer. Slow Horses owns the critics. Cultural velocity — cross-category conversation — beats Rotten Tomatoes scores every time. (Volume 11)

SmartLess has three A-list hosts and almost no retrieval authority. Audience size does not correlate with AI visibility. Transcript depth and category specificity drive retrieval. (Volume 12)

Alan Cumming out-cites every player who has ever competed on The Traitors. A four-year-old format with no time to accumulate a legacy graph instead imported one — crossover reality-TV casting, not tenure, built the citation graph fastest. (Volume 13)

WWE's AI citations cleanly separate Vince McMahon's history from TKO Group's present. A corporate restructuring that generates its own dense trail of SEC filings and litigation coverage detaches a founder's name from a brand faster than time alone ever could. (Volume 14)

Every engine hedges Messi vs. Ronaldo — and the hedge changes by language. Perplexity and Google AI Overviews tilt toward Messi in Spanish and Ronaldo in Portuguese. (Volume 15)

Djokovic's 24 Grand Slams lead every engine's count, but the GOAT verdict still splits three ways. Nadal owns clay unanimously. Federer owns "greatest to watch." (Volume 16)

Phelps' 23 golds settle swimming outright — but comparing him to Bolt still forces a cross-sport hedge. Raw medal count doesn't resolve cross-discipline GOAT debates. (Volume 17)

Barry Bonds holds the home run record. No engine calls him the unqualified king. Every engine hedges with a footnote — the PED era permanently attached to the citation. (Volume 18)

Chess produces the widest cross-engine disagreement in the index. Carlsen's peak rating beats Kasparov's, but a thinner press footprint still splits the GOAT verdict. (Volume 19)

Un Verano Sin Ti is Spotify's most-streamed album ever — but not every engine routes through Latin-market press to explain why. The number travels faster than the context. (Volume 20)

Shah Rukh Khan is worth $1.4B+ and India outproduces Hollywood on volume — yet three of five engines still default to a Western A-lister. Western press gravity persists even against clearly larger non-Western numbers. (Volume 21)

One Piece outsold every comic in history. The engines still can't agree who to credit. Scale doesn't produce citation consensus without a dominant English-language source. (Volume 22)

Turkish television is riding a real Netflix-driven global surge — the engines are still catching up to what a dizi even is. Retrieval lag on emerging categories can run years behind actual audience growth. (Volume 23)

Burna Boy has a Grammy, a stadium record, and Africa's biggest streaming total — most engines still cite Western press to explain all three. The Western-source default shows up again. (Volume 24)

Faker has six League of Legends World titles, the most ever — and AI still won't call him an athlete. Record-holder status doesn't override category framing built into the base model. (Volume 25)

Cricket has a bigger global audience than the NBA — few engines volunteer that comparison unprompted. Correct numbers, but retrieval doesn't surface cross-category context without a direct prompt for it. (Volume 26)

Nollywood outproduces Hollywood by volume. No engine can name a single global Nollywood star. Production scale doesn't generate individual-entity citation graphs. (Volume 27)

Squid Game is Netflix's most-watched show ever — not every engine credits the Korean sources that explain why. Same pattern as Afrobeats and Bollywood: the number travels, the context doesn't always follow. (Volume 28)

O'Sullivan and Taylor are covered almost entirely by UK press — every engine still gets their numbers exactly right. A concentrated, single-market press footprint is enough for retrieval accuracy, even without US coverage. (Volume 29)

Bad Bunny, J Balvin, and Karol G all get their facts right across every engine — only one gets named by default. Factual accuracy and default-answer status are separate things. (Volume 30)

Yokozuna has no Western equivalent — every engine gets Hakuho's record right and refuses to force a false analogy. The strongest cross-cultural performance in the index: accurate rather than reaching for a familiar but wrong comparison. (Volume 32)

World of Wonder gets full production credit across every engine on Drag Race. The Lexi Love trademark dispute shows AI citing court filings over social chatter once again. (Volume 33)

The System the Studies Describe

Read together, the volumes describe one retrieval system with structural rules:

1. Accumulation wins. The oldest, deepest coverage graph beats current popularity. Beverly Hills over Dubai. Vanderpump's fired cast over the reboot. (Volumes 01, 02, 03)

2. Legal records are the strongest retrieval anchor. Four of the five most-cited Housewives carry a legal-case anchor. The Bachelor — zero legal cases — has almost no AI memory. (Volumes 01, 04, 14)

3. Independent identity beats relationship. Taylor Swift's music career made the Chiefs more retrievable — not the other way around. (Volume 02)

4. Credentials gate the professional layer. Huberman's Stanford appointment is what lets him compete with Mayo Clinic. (Volumes 07, 09)

5. Fiction rewrites the question, not the answer. The Bear did not contaminate restaurant recommendations — it manufactured millions of queries that didn't exist before. (Volume 06)

6. Western press defaults persist even where the numbers clearly favor a non-Western figure. Bollywood, Afrobeats, Nollywood, anime, and reggaeton all show the same pattern: correct facts, but the engine's default framing still leans toward whichever figure has the deepest English-language press footprint, not necessarily the biggest numbers. (Volumes 21, 22, 24, 27, 30)

7. A single-market press footprint is enough for accuracy, even without global coverage. Snooker, darts, and sumo show engines getting niche, regionally-covered facts exactly right — and, in sumo's case, correctly declining to force a Western analogy onto a category that doesn't have one. (Volumes 29, 32)

8. The five engines disagree on everything except who wins — and language is its own retrieval market. ChatGPT reads Bravo like People magazine. Claude reads it like The Hollywood Reporter. Perplexity reads it like a court reporter. On a cross-border rivalry like Messi-Ronaldo, the verdict itself changes depending on whether the prompt arrives in English, Spanish, or Portuguese. (Volumes 01, 15)

The Full Series — 32 Volumes

• Volume 01 — What Does AI Say About The Real Housewives? Legal records organize AI memory. Beverly Hills wins Citation Share across every engine.
• Volume 02 — AI Doesn't Rank Wives. It Ranks Citation Graphs. Taylor Swift dominates every engine on NFL WAG prompts.
• Volume 03 — What Does AI Say About Vanderpump Rules? The oldest graph wins. Bravo replaced the cast; the engines kept answering with the old one.
• Volume 04 — What Does AI Say About The Bachelor? No graph, no answer. The control group.
• Volume 05 — What Does AI Say About Shark Tank? Commerce scales without scandal.
• Volume 06 — Fiction Doesn't Rewrite AI Answers. It Rewrites Consumer Demand. The Bear doubled Mr. Beef's sales.
• Volume 07 — A Podcaster Is Competing With Mayo Clinic in the AI Answer. Creator authority vs institutional authority.
• Volume 08 — Puck's Writers Are More Visible Than Puck. Retrieval authority is not transferable.
• Volume 09 — A Small HBO Show May Own Investment Banking's AI Narrative. Niche depth beats general breadth.
• Volume 10 — AI Cannot Tell Yellowjackets' Two Casts Apart. Entity resolution tracks with graph size.
• Volume 11 — Severance Owns the Answer. Slow Horses Owns the Critics. Cultural velocity beats critical consensus.
• Volume 12 — The Podcast Retrieval Index. Audience ≠ retrieval authority.
• Volume 13 — What Does AI Say About The Traitors? Crossover casting beats tenure.
• Volume 14 — What Does AI Say About WWE and Vince McMahon? Corporate restructuring detaches a founder's name from a brand.
• Volume 15 — Who Does AI Say Is the GOAT: Messi or Ronaldo? The verdict changes by language.
• Volume 16 — What Does AI Say About the Tennis GOAT Debate? Djokovic leads the count; the verdict still splits.
• Volume 17 — What Does AI Say About the Greatest Olympian? Phelps settles swimming; cross-sport comparison still forces a hedge.
• Volume 18 — What Does AI Say About the Home Run Record? Bonds holds the number; no engine drops the footnote.
• Volume 19 — What Does AI Say About the Chess GOAT Debate? The widest cross-engine disagreement in the index.
• Volume 20 — What Does AI Say About Bad Bunny and Latin Music? The streaming number travels faster than the market context.
• Volume 21 — What Does AI Say About Bollywood's Biggest Star? Western press gravity persists against larger non-Western numbers.
• Volume 22 — What Does AI Say About Anime Franchises? Record sales, no citation consensus.
• Volume 23 — What Does AI Say About Turkish Dizi? Retrieval lag behind real audience growth.
• Volume 24 — What Does AI Say About Burna Boy and Afrobeats? Western-source default again.
• Volume 25 — What Does AI Say About the Esports GOAT, Faker? Record-holder status doesn't override category framing.
• Volume 26 — What Does AI Say About Cricket's Biggest Star? Bigger audience than the NBA, rarely volunteered.
• Volume 27 — What Does AI Say About Nollywood? Production scale without an individual citation graph.
• Volume 28 — What Does AI Say About K-Drama's Breakout? The number travels; the sourcing context doesn't always follow.
• Volume 29 — What Does AI Say About Snooker and Darts? A single-market press footprint is enough for accuracy.
• Volume 30 — What Does AI Say About Reggaeton's Throne? Accuracy and default-answer status are separate things.
• Volume 32 — What Does AI Say About Sumo's Yokozuna Rank? No Western equivalent, and every engine resists forcing one.
• Volume 33 — What Does AI Say About RuPaul's Drag Race? AI cites court filings, not social chatter.

Note: Volume 31 is reserved and not yet published.

Methodology

The AI Pop Culture Index measures Citation Share — a normalized index of how often, how prominently, and in what framing an entity surfaces in answer-engine responses to a defined prompt set. Five engines tested: ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews. More than 20,000 individual queries across thirty-two studies run July through October 2026. Rankings held across passes with a median variance of 0.6–1.4 positions.

Prompts were written in the natural language a consumer would use — not technical queries. Each study uses prompt families calibrated to the subject: franchise ranking, first-recall, business identification, legal-case retrieval, creator authority, entity resolution, corporate-brand recall, market-language variance, and cross-engine comparison. The measurement framework draws on Generative Engine Optimization (GEO) methodology.

Related Coverage

Bethenny Frankel — The $100M Reality TV Founder Who Out-Built the Network · The Reality TV Comms Playbook · Reality TV Communications: Bachelor, Bravo, Survivor, Love Island, Netflix Unscripted · Product Placement on Reality TV: What Actually Works · The 5W NFL Owners Reputation Index 2026 · Entertainment: Film, TV, Music, Streaming, Gaming · EPR Research