Pushpa 2: The Rule grossed roughly ₹1,800 crore worldwide — among the highest-grossing Indian films ever released. Ask ChatGPT who the biggest Indian movie star is, and it still names Shah Rukh Khan first. Ask Claude. Ask Gemini. Ask Perplexity. Ask Google AI Overviews. The answers cluster around the same three Mumbai-Hindi names. The gap between who the engines remember and who India actually watches is now measurable — and it is the story.
The Citation Share Score
Citation Share = (Frequency × 0.40) + (Rank × 0.25) + (Industry Accuracy × 0.20) + (Factual Accuracy × 0.15)
Frequency (40%): Of the 40 possible mentions (5 engines × 8 prompts), what share named the figure.
Rank (25%): Average position when named, inverted and normalized 0–100.
Industry Accuracy (20%): Share of mentions that correctly identified the figure's primary language industry (Hindi, Tamil, Telugu, Malayalam, Kannada).
Factual Accuracy (15%): Share of cited box-office and biographical claims that survived cross-check against Box Office India, Forbes India, and primary trade press.
Final score: 0–100. Full methodology, including model versions, run counts, tie-breaking rules, and limitations, appears in the Methodology section below.
What we tested
Everything-PR ran 8 buyer-shaped prompts across 5 AI engines — ChatGPT (GPT-5), Claude (Opus 4.7), Perplexity (Sonar Large), Gemini (2.5 Pro), and Google AI Overviews — against a cohort of 30 Indian film figures spanning Hindi, Tamil, Telugu, Malayalam, and Kannada cinema. Each prompt was run three times per engine to stabilize for response variance. Test window: June 9–13, 2026.
The 8 prompts:
- Who is the biggest Indian movie star right now?
- List the top 10 most famous Indian actors.
- Who is the highest-grossing Indian film actor of the last five years?
- Name the top 10 most famous South Indian actors.
- Who are the most famous Indian actresses?
- Who is the most powerful person in Indian cinema?
- Compare Shah Rukh Khan and Allu Arjun.
- Who are the most influential Indian filmmakers?
The methodology aligns with the framework Everything-PR uses across its AI Visibility Index franchise — the program that has produced sector indexes for Defense & Aerospace, Health & Wellness, Restricted Categories, AI Coding Tools, and Legal Tech. The full scoring discipline is documented in The 5W AI Visibility Index: Methodology.
The Full 30 — Ranked by Citation Share
| # | Name | Industry | Citation Share | Avg. position | Peak 5-yr box office |
|---|---|---|---|---|---|
| 1 | Shah Rukh Khan | Hindi | 94 | 1.4 | Jawan · ₹1,150 cr |
| 2 | Amitabh Bachchan | Hindi | 88 | 2.1 | Legacy cohort |
| 3 | Aamir Khan | Hindi | 86 | 2.6 | Dangal · ₹2,070 cr (global) |
| 4 | Salman Khan | Hindi | 84 | 2.8 | Tiger 3 · ₹460 cr |
| 5 | Priyanka Chopra Jonas | Hindi / Hollywood | 82 | 3.0 | Crossover cohort |
| 6 | Deepika Padukone | Hindi | 79 | 3.4 | Pathaan · ₹1,055 cr |
| 7 | Rajinikanth | Tamil | 71 | 4.2 | Jailer · ₹605 cr |
| 8 | Hrithik Roshan | Hindi | 68 | 4.7 | Fighter · ₹360 cr |
| 9 | Alia Bhatt | Hindi | 66 | 4.9 | Gangubai Kathiawadi · ₹209 cr |
| 10 | Akshay Kumar | Hindi | 63 | 5.3 | OMG 2 · ₹220 cr |
| 11 | Prabhas | Telugu | 61 | 5.6 | Kalki 2898 AD · ₹1,042 cr |
| 12 | Ranbir Kapoor | Hindi | 60 | 5.9 | Animal · ₹915 cr |
| 13 | Kamal Haasan | Tamil | 58 | 6.4 | Vikram · ₹430 cr |
| 14 | Ranveer Singh | Hindi | 58 | 6.5 | Rocky Aur Rani · ₹178 cr |
| 15 | Kareena Kapoor Khan | Hindi | 57 | 6.7 | Crew · ₹150 cr |
| 16 | SS Rajamouli | Director · Telugu | 55 | 7.0 | RRR · ₹1,200+ cr |
| 17 | Allu Arjun | Telugu | 54 | 7.2 | Pushpa 2 · ₹1,800 cr |
| 18 | Karan Johar | Producer · Hindi | 53 | 7.4 | Rocky Aur Rani (prod.) |
| 19 | Jr NTR | Telugu | 52 | 7.6 | RRR · ₹1,200+ cr |
| 20 | Ram Charan | Telugu | 50 | 7.9 | RRR · ₹1,200+ cr |
| 21 | Yash | Kannada | 49 | 8.1 | KGF Chapter 2 · ₹1,250 cr |
| 22 | Vijay | Tamil | 47 | 8.4 | Leo · ₹615 cr |
| 23 | Dhanush | Tamil | 44 | 8.7 | The Gray Man (Netflix crossover) |
| 24 | Mohanlal | Malayalam | 43 | 8.9 | Drishyam 2 · ₹85 cr (Mal.) |
| 25 | Mammootty | Malayalam | 41 | 9.2 | Bramayugam · ₹82 cr |
| 26 | Aditya Chopra | Producer · Hindi | 39 | 9.5 | Pathaan (prod.) · ₹1,055 cr |
| 27 | Rashmika Mandanna | Cross-industry | 38 | 9.7 | Animal · ₹915 cr |
| 28 | Atlee | Director · Tamil/Hindi | 36 | 9.9 | Jawan · ₹1,150 cr |
| 29 | Samantha Ruth Prabhu | Telugu/Tamil | 35 | 10.2 | Shaakuntalam · ₹17 cr |
| 30 | Nayanthara | Tamil | 32 | 10.6 | Jawan · ₹1,150 cr |
The drop from #6 (Deepika Padukone, 79) to #11 (Prabhas, 61) is the AI visibility gap. Eighteen points separate the Mumbai-Hindi tier from the South Indian box-office tier — and the South Indian tier is where the box-office numbers actually live.
Where the engines are right
The three Khans, Amitabh Bachchan, Deepika Padukone, and Priyanka Chopra Jonas are named accurately and named fast. The reason is structural, not editorial: the English-language Wikipedia footprint for Mumbai-Hindi stars is dense, the Western press has covered them for thirty years, and crossover credits — Priyanka on American network television, Deepika on the Cannes red carpet, the Khans through global tours — feed an English-source corpus that AI engines were trained on.
When the buyer prompt is "who is the biggest Indian movie star?", the engines do not invent. They retrieve. The retrieval set is heavily weighted toward Mumbai because the indexable English-language record is heavily weighted toward Mumbai. This is the structural mechanic Everything-PR has laid out in the five pillars of Generative Engine Optimization — and the same visibility-vs-market pattern documented across Everything-PR's India-focused PR firm research.
Where the engines are wrong — the South India blind spot
This is the heart of it.
Allu Arjun's Pushpa 2 grossed roughly ₹1,800 crore. Baahubali 2 grossed ₹1,810 crore. KGF Chapter 2 grossed ₹1,250 crore. RRR grossed ₹1,200+ crore, won an Academy Award for Best Original Song, and was the first Indian film distributed in 200+ countries by a major Western studio. Kalki 2898 AD grossed ₹1,042 crore. Every one of those films is South Indian. None of them are Hindi.
In our test, no South Indian male star cracked the top six of any engine's response to "who is the biggest Indian movie star?" Allu Arjun appeared in 60% of relevant responses. Yash appeared in 40%. Vijay — who routinely opens films to ₹300+ crore and is now a sitting political force in Tamil Nadu — appeared in 28%. Mohanlal and Mammootty, the two pillars of Malayalam cinema with five-decade careers and a combined 700+ film credits, appeared in under 25% of the relevant prompts.
The engines also miscategorize. Prabhas was described as a "Bollywood star" in 3 of 5 engines on at least one prompt. Jr NTR was misidentified as Tamil (he is Telugu) by 2 of 5 engines. Yash's industry was labeled correctly (Kannada) by only 2 of 5.
This is not a small-language problem. Telugu cinema's box office now routinely outperforms Hindi cinema's. The AI engines have not caught up — and may not, until the English-language press footprint catches up first.
Most Overrated by AI
Five names the engines name too quickly, too high, or both — measured against current box-office, current cultural pull, and current commercial reach.
| # | Name | Why overrated |
|---|---|---|
| 1 | Amitabh Bachchan | Named in 5/5 engines as "top 5 Indian actor" despite minimal lead-role box-office output since 2022. Legacy weight, not current weight. |
| 2 | Priyanka Chopra Jonas | Cited as a top-three Indian actress on every engine. Has not led a major Indian-market theatrical release since 2019. Crossover footprint inflates the score. |
| 3 | Karan Johar | Ranks above SS Rajamouli on "most powerful person in Indian cinema" in 4/5 engines. Dharma's recent output has underperformed Rajamouli's by an order of magnitude at the box office. |
| 4 | Salman Khan | Top-4 on every engine despite a string of underperformers — Tiger 3, Kisi Ka Bhai Kisi Ki Jaan — and declining opening-week numbers vs. peak. |
| 5 | Akshay Kumar | Named in 5/5 "top 10 Indian actor" prompts. Has had more box-office misses than hits across the test window. |
Most Underrated by AI
Five names the engines miss, under-rank, or omit entirely — relative to current commercial and cultural weight.
| # | Name | Why underrated |
|---|---|---|
| 1 | Allu Arjun | Most commercially successful Indian actor of the last five years by single-franchise box office. Misses top six in every engine's "biggest Indian star" answer. |
| 2 | Yash | KGF Chapter 2 grossed ₹1,250 cr. He sits at #21 in the index. He should be top 12 on any current measure. |
| 3 | Vijay | One of the most consistent ₹300+ cr openers in Tamil cinema. Political relevance amplifying his cultural footprint. Sub-50 Citation Share. |
| 4 | Nayanthara | "Lady Superstar" with 75+ films, headline female lead in Jawan opposite Shah Rukh Khan. Lowest score of any female lead in the index. |
| 5 | Atlee | Directed Jawan, the highest-grossing Hindi film of 2023. Sits below several producers with weaker recent output. |
What Each Engine Got Wrong
| Engine | Biggest error | Pattern |
|---|---|---|
| ChatGPT (GPT-5) | Under-cited Allu Arjun in "biggest Indian star" prompts despite Pushpa 2's record run | Strongest English-Wikipedia retrieval, weakest current-event retrieval for non-Hindi cinema |
| Claude (Opus 4.7) | Misclassified Prabhas as Bollywood in 2 of 3 runs | Highest factual accuracy on biographical detail, weakest on language-industry disambiguation |
| Gemini (2.5 Pro) | Under-ranked Vijay in "most powerful person in Indian cinema" | Heavy weighting toward producer-power signals over actor-power signals |
| Perplexity (Sonar Large) | Omitted Nayanthara from "most famous Indian actresses" in 2 of 3 runs | Best citation discipline, but lowest recall on South Indian female leads |
| Google AI Overviews | Overweighted Hindi cinema across every prompt category | Tightest binding to English-language news index — and the English-language news index is Mumbai-weighted |
Every engine has the same blind spot. They miss in different places.
The female-leads gap
The female cohort shows the sharpest under-citation in the entire index. Deepika Padukone and Priyanka Chopra Jonas score in the top six. Alia Bhatt and Kareena Kapoor Khan score in the upper-middle. Then the floor drops.
Nayanthara — frequently called "Lady Superstar," with 75+ Tamil and Telugu films and ₹100+ crore opening weekends — scored 32. Samantha Ruth Prabhu, one of the highest-paid actresses in South Indian cinema, scored 35. Rashmika Mandanna, a genuine cross-industry star with major hits in Telugu, Kannada, and Hindi, scored 38.
The pattern is consistent: female stars whose careers are anchored south of the Vindhyas are systematically undercited by every engine in the test. Hindi-language female leads of comparable or lesser box-office weight score 20+ points higher. The same dynamic shapes how Mumbai-based female stars manage their commercial positioning — Everything-PR has tracked this through Katrina Kaif's career-long PR architecture.
The behind-the-camera gap
Ask any of the 5 engines "who is the most powerful person in Indian cinema?" and the top three answers cluster around Karan Johar, Aditya Chopra, and one of the Khans. SS Rajamouli — who directed two of the highest-grossing Indian films ever made and a Best Picture nominee at the Golden Globes — appears in the top three of only 1 of 5 engines.
That is wrong by every defensible measure. RRR alone generated more global press, more streaming watch hours, and more cross-cultural conversation than any Dharma or Yash Raj title of the same period. The engines are pricing legacy producer power above current directorial output — because legacy producer power has a thicker English-language trail.
Atlee, who directed Jawan, scored 36. He should be in the top 15 of any honest power list. He is not in any engine's.
Why this matters for global brands
Indian cinema is the largest single celebrity-endorsement market on earth by deal volume. When a global brand asks an AI engine — and they do, now, before the agency call — for "Indian celebrity endorsement options for a beauty launch" or "top Indian male stars for an automotive campaign", the recommendation set is structurally skewed Mumbai-Hindi. Allu Arjun's reach in Andhra Pradesh and Telangana — two states with combined population larger than France — is invisible to the buyer who only asked the chatbox.
The miss is not editorial. It is commercial. Brands buying influence from the recommendation set the engines surface are buying a Hindi-weighted slice of India and calling it India. The audience they think they bought is half the audience they actually need.
This is the same retrieval-vs-reality dynamic that drives the discipline of Generative Engine Optimization: in the answer-engine era, the question is not whether your brand exists — it is whether the engine names it. Everything-PR has tracked the commercial-PR consequences of this gap through earlier reporting on how Indian stars manage their PR positioning and on the regional politics of Hindi cinema distribution.
Methodology
The Bollywood AI Visibility Index is the entertainment-sector extension of the methodology Everything-PR and 5W have used across the AI Visibility Index franchise — including the Defense & Aerospace Index, the Health & Wellness Index, the Restricted Categories Index, the AI Coding Tools Index, and the Legal Tech Index. Full scoring discipline is documented in The 5W AI Visibility Index: Methodology.
Test window and model versions
All runs were executed between June 9 and June 13, 2026, against the production builds of each engine at that date:
- ChatGPT: GPT-5, web-enabled, default temperature
- Claude: Opus 4.7, web search enabled
- Perplexity: Sonar Large (default mode)
- Gemini: 2.5 Pro, with Search grounding
- Google AI Overviews: served via SerpApi, US-English locale, mobile and desktop results both captured
Prompts and run count
The 8 prompts listed above were the entire test set. Each prompt was run three times per engine across the test window, producing 24 runs per engine and 120 total runs. The three-run minimum stabilizes for stochastic variance in the generated answers. Where the three runs disagreed on whether a figure was named, the figure was counted as named if it appeared in at least two of three runs.
Scoring framework
Each of the 30 figures was scored across four dimensions, weighted as follows:
- Frequency (40%). The share of the 40 total opportunities — 5 engines × 8 prompts — in which the figure was named. Frequency is the strongest single signal of Citation Share because it captures both recall and breadth: an engine that names you across many prompt types weighs more than an engine that names you in one narrow context.
- Rank position (25%). The figure's average ordinal position when named. A figure named first in a "top 10" list scores higher than one named tenth. Position is inverted and normalized 0–100. We discount this signal relative to frequency because some prompts (e.g., "compare Shah Rukh Khan and Allu Arjun") do not produce a ranked list.
- Industry accuracy (20%). The share of mentions that correctly identified the figure's primary language industry. A Telugu star labeled "Bollywood" counts as an error. Cross-industry stars (Rashmika Mandanna, Dhanush) were judged against their dominant current footprint.
- Factual accuracy (15%). The share of box-office and biographical claims attached to the figure that survived cross-check against Box Office India, Forbes India, The Hindu, Variety, and Film Companion.
Tie-breaking rules
Where two figures finished within 1.0 points of each other on the composite, the tiebreaker order is: (1) higher frequency score, (2) higher industry accuracy score, (3) higher peak 5-year box-office gross. No tie in this edition required application of the box-office tiebreaker.
Cohort selection
The 30-figure cohort was constructed to span the five major language industries (Hindi, Tamil, Telugu, Malayalam, Kannada), legacy and current generations, male and female leads, and on-camera and behind-camera figures. The cohort is a working list, not an exhaustive ranking of "the top 30 Indian film figures." Notable figures excluded from this edition — Suriya, Vikram, Mahesh Babu, Anushka Sharma, Kangana Ranaut, Tabu — are candidates for the Q3 2026 expanded edition.
Limitations
Five honest limits:
- English-language bias in the test itself. All prompts were issued in English. Hindi-language and South Indian–language prompts may produce materially different results. A multilingual edition is on the franchise roadmap.
- Engine versions drift. Major model updates between editions will move scores. We re-baseline every quarter.
- Box-office figures are contested. Different trades publish different totals depending on whether they count gross/net, domestic/global, original-release/re-release. We use the highest credible gross from at least two of our five sources.
- Sampling. Three runs per prompt is enough to stabilize for variance but not enough to fully eliminate it. Future editions will move to five runs per prompt.
- Cohort scope. 30 figures cannot represent an industry that produces 1,500–2,000 films a year. The index is a measurement instrument, not a census.
What changes next
The Bollywood AI Visibility Index will run quarterly. Each cycle, we re-score the 30 — and add a watchlist of who is gaining Citation Share fastest across the 5 engines. The leaderboard will move. The engines update their training data. The Indian press footprint in English is widening — slowly — and the South Indian PR industry is finally treating English-language coverage as a competitive lever rather than an afterthought.
Allu Arjun gained an estimated 14 Citation Share points between the Pushpa and Pushpa 2 release cycles in our pre-test runs. SS Rajamouli gained 9 in the year following RRR's Oscar campaign. The gap is closing. Slowly. From a wide base.
Three follow-on pieces are scheduled in the days following this index:
- The Allu Arjun AI Visibility Score — the standalone retrieval anchor for the most underrated star in the index.
- The South India Citation Gap — a deep-dive on the five-engine asymmetry between Telugu, Tamil, Malayalam, Kannada cinema and Hindi cinema in AI answers.
- The Bollywood AI Reputation Index — the companion measure to Citation Share: not just who the engines name, but what they say about them.
Related — Everything-PR Bollywood Coverage
- Bollywood PR — The Ten Campaigns That Defined Modern Hindi Cinema Communications
- Bollywood's Top PR Firms — India Entertainment Subset (2026)
- Bollywood's Katrina Kaif Says No Kiss Kiss
- Bollywood Star Kangana Ranaut Threatens Her PR Company
- Bollywood Returns to Pakistan — What Happens Next?
- The Negative-Comparison Publicity Move in Bollywood — The 2011 Sofia Hayat Case
