Stringers
Mobile App · Android · written 11-17
Aramızda: Flört & Arkadaşlık · Appion Digital LLc

After 1,827 installs it reached the Turkey dating Top Grossing chart. The store page says "goodbye fake profiles," and the terms of service say "includes AI virtual profiles."

What it is A Turkish-language local dating app: it finds people nearby by neighborhood and on a map, has six-player bottle-spin-style game rooms, and makes money through Gold memberships and coins. Its terms of service state that the platform includes AI-managed virtual profiles that chat with you.

Why it took off It hasn't really taken off yet: the bar to reach the Turkey dating Top Grossing chart is low, and a thousand-plus installs plus a few paying users can get it there. The money comes from a paywall plus AI chat companions that make male users feel "someone is replying," with ads that feed payment data back for optimization.

CHow you could follow Don't copy the Turkey reskin wave; learn the playbook: in smaller-language regions, find dating or social Top Grossing charts with only dozens of entries, then use clearly labeled AI companion chat or gamified social features with a membership paywall, and buy a small amount of installs to validate paying users.

Grade C: it hasn't broken out itself, Turkey already has a dozen-plus similar apps, and the established ones have two million installs; the core revenue mechanic is AI posing as real people, which platforms are now watching closely after 2026-09. What's worth learning is the low-barrier chart and reskin-across-regions playbook; copying the app itself isn't worth it. Few people know about it yet

Est. monthly revenue (Google Play) under $1KWe estimated this from its rank on country top-grossing charts and real install counts, calibrated against real revenue; typical error is 2–3x. Read it as a tier, not an exact number.

Data (updated daily) · first launch 2026-10-16 · 11-20 installs 3,110
Android · Total installs worldwide
1k2k3k11-1611-1711-1811-1911-203.1k
Data 2026-11-16 → 2026-11-20 (5 days, 5 observations). Google's detail page only refreshes install counts every other day, so a flat day doesn't mean nobody installed.

Aramızda is a Turkish-language local dating app. It launched on Oct 16, got 1,827 installs in a month, and in Nov 18 it was ranked #49 on the Turkey Google Play dating Top Grossing chart. The rank sounds good, but that chart only had 74 apps that day, so it hasn't really broken out. What's worth looking at is the revenue combo: memberships and coins as the paywall, AI-played "users" chatting with male users, plus paid acquisition with payment data fed back. Dozens of teams in Turkey have been running this for 2026 years, so the value of following is mainly in learning the playbook, not copying this app.

The store page and the terms of service say two different things

On the Turkey storefront of Aramızda, the Safety section says:

Mavi Tikli Doğrulanmış Profil: Selfie doğrulama teknolojisi ile sahte profillere son ver, gerçek kişilerle eşleş. (Blue-check verified profiles: end fake profiles with selfie verification, match with real people.)

Below that is a line "%100 güvendesin," meaning "you are 100% safe."

The terms of service linked at the bottom of the store page (sparkin.online, signed "last updated 2026 Aug 11"), section 3 is titled "AI-supported virtual profiles":

The platform may include virtual profiles managed by artificial intelligence... they do not represent real, verified people, and automated replies generated by AI models sustain the chat. By continuing to use this app, you accept that virtual profiles may exist on the platform and that they are not real users.

The privacy policy (Oct 17) is more specific: conversations with these profiles are "sent via OpenRouter to the relevant LLM provider" to generate replies. The official homepage also says "the platform may include AI-powered virtual profiles."

That's the business. The store page is written for users; the terms are written for reviewers and lawyers. The terms are dated two months before the app launched, which suggests a ready-made template (inferred). Swap the name and you can launch another app.

How it grew

DateTotal installsTurkey Dating Top GrossingNigeria Dating Top Grossing
11-161,02755171
11-171,02752163
11-181,82749153

(Google Play installs and charts, 11-16 to 11-18)

A month after launch, it averages about thirty-odd installs a day; third-party estimates put it around 90 a day in mid-November. One day it jumped by 11-18 (800). That could be paid acquisition ramping up, or Play doing a one-time install count update. From outside, you can't tell. It never charted on any Top Free list, only Top Grossing. Few people install it, but some people pay.

Take the rank with a grain of salt. 11-18 The Turkish dating Top Grossing chart has only 74 entries: #23 Gönlünegöre has just 630 installs, #48 Datte has 6,511, and #61 Denk has 1,497. On a chart this thin, a thousand-odd installs plus a few payers is enough to land a spot. So "#49" only tells you the app is taking in some money. How much is unclear, and it's probably not much (inference).

Where the money comes from: Gold, coins, and people who always reply

The paywall is all in the store listing. Aramızda Gold unlocks: seeing who liked you and who viewed you, unlimited likes, unlimited unsends, Boost (getting shown to more people in your area), and invisible browsing. Sending gifts, voice calls, and video calls also cost Gold or coins. Subscriptions run on Google Play auto-renewal.

Turkey's dating apps have more men than women. Male users download the app, like profiles, and if nobody replies for a while, they quickly uninstall. The AI profiles keep them feeling someone is replying, and the warmer the replies, the more likely they are to tap "who liked you" or buy coins for video calls. Turkey is already a high-ground for paid social dating: in June 2024, Turkey's top-ten app revenue included Azar, Tinder, Tango, LivU, and WePlay, which together made over $3M that month (ensonhaber, 2024-07-03).

There's no technical barrier. The privacy policy lists the whole stack: Firebase for login and push, Cloudflare R2 for photo storage, Agora for audio and video, OpenRouter for LLM access, and Google Play for payments.

Where users come from: most likely paid

Nobody online is talking about it. Nobody has mentioned it on X in nearly 30 days. A web search only turned up an iOS dating-coach app with the same name from a different developer. No press coverage, no creators, and not a single review on the store page yet.

Paid acquisition (inferred) is the most likely explanation for its installs and payments. Three pieces of evidence:

  • The privacy policy states that data goes to "Meta (Facebook) / Google Analytics, used to measure ad conversions." Only apps that run ads optimized for payments need to send conversion data back.
  • The developer, Appion Digital LLC, is a verified US advertiser in Google's Ads Transparency Center, with about 3 ads. The page notes that "some ads contain age-restricted content," and the creatives can't be viewed without logging in.
  • The install curve doesn't look like organic growth. Fubby, from the same wave, has 229 ads in the TikTok Ad Library. This batch of apps grew by buying traffic.

Searching Meta's Ad Library by app name, package name, or the store-page tagline "Semtindeki insanlar burada" turned up nothing. Ad spend may still be small, or they may be using a different page name.

Two smaller inputs: the title "Flört & Arkadaşlık" (Dating & Friendship) lines up with the two biggest search terms in the Turkish store. It also appears on Nigeria's Dating Top Grossing chart, probably squeezed into the long tail with a bit of paid spend.

On the same chart, a dozen nearly identical apps

Aramızda isn't a pioneer. It's a late entrant in a wave of reskinned apps. 11-18 On the Turkish dating Top Grossing chart, several same-type "Turkish-language chat + paid coins" apps (Google Play installs, 11-18):

ProductLaunchedTotal installsTurkey Dating Top GrossingNotes
Civarda Sevgili Bul2021-112.3M4An old "people nearby" coin-chat app, with a pile of fake-profile and surprise-charge complaints on the complaints site Şikayetvar
Kaçamak Sevgili Bul2022-091.61M9Veteran in the same category
Fubby2026-0849k21Turkish team, with a Missouri, US LLC as the entity, 229 TikTok ads
Hovarda / Datte2026-05 / 0823k / 6.5K22 / 48One developer, two brands
Alev / Çapkın2026-04 / 0835k / 12k35 / 44Two brands under one LLC
Yakınlarda2025-0878k55Also "see people nearby on a map"; the concept is closest to Aramızda's
Aramızda2026-101,82749The subject of this article

The chart also has Pawn, Tenhada, KiDate, Bontik, Gönlünegöre, and Denk, which entered between autumn 2025 and October 2026, with installs ranging from a few hundred to twenty thousand. At the top are the real-person giants: Tinder (#1), Bumble (#3), and Badoo (#7).

This batch follows a very consistent playbook: a US LLC as the developer entity, Turkish-language UI and terms, Turkish law as the governing law, the same paywall, and one team running two or three brands. Aramızda's developer account has only this one app, and it's unclear who is behind it. The website, terms, and personal-data compliance text are all in Turkish, which suggests a Turkish team using a US LLC as a shell (inferred).

Neighborhoods, maps, six-person game rooms: none of these explain why it charted

Aramızda sets itself apart from its peers on the store page in three ways:

  • Local matching down to province, district, and neighborhood (il / ilçe / semt), with "Semtindeki insanlar burada" (the people in your neighborhood are here). But the privacy policy says it collects "city-level location," which doesn't match the store page.
  • A map mode, plus a Ghost mode for going invisible at any time.
  • Three real-time six-player game rooms (Spin the Bottle, Hot Seat, and Blind Match), plus two-player mini-games in chat: This or That, compatibility tests, emoji duels, and truth or dare.

The veteran of the bottle-spinning mechanic is Spin the Bottle, launched in 2016, with 13.6M lifetime installs. Putting it inside a dating app is fresh for Turkey's coin-based chat.

But with a thousand-plus installs, real people can't fill six-person rooms. The rooms are either empty or depend on AI profiles to fill them (inferred). It got onto the Top Grossing chart through the same combination as Civarda and Fubby. The game rooms feel more like hooks for ad creatives and store screenshots.

If it were you

C: The playbook is worth learning, but don't follow it into Turkey. It hasn't proven itself yet. The same chart, which ranks #74, already has a dozen nearly identical apps, and the established ones have two million installs. Its paid model rests on AI impersonating real people. In September 2026, Anthropic disclosed that one developer had used 20 AI personas across 4,700 dating apps to chat and sell coins. Apple and Google were both notified. Since then, that exact practice has been a priority for platform enforcement. The store page says "goodbye fake profiles," and the terms say "AI profiles are included." That's a ready-made reason for removal.

Who it's for: small teams that can run Meta / Google app-install ads and work out whether first purchases pay back; teams that already have a dating or social backend and can reskin it into a new language version in a few weeks; ideally, teams with an overseas company entity that can be verified as an advertiser.

Who it's not for: anyone who can't run paid ads and is hoping for organic store traffic (apps like this have no organic reach at all); anyone unwilling to touch the grey area of "AI chat passed off as real people" who still wants to earn the same way; anyone building a long-term brand and trying to earn reputation (look at Civarda's complaints page to see what the reviews will look like).

How to do it: if you build this, build the half it gets wrong, the half others can't easily copy: put the AI on the table. Make an openly AI chat or "practice flirting" app, or add an openly labeled AI character to a real-person social app (for icebreakers, or to play game rooms with you), and copy Gold's paywall structure. Don't pick Turkey as the market. Look for smaller-language markets that haven't been filled by this wave of reskins.

How to acquire users: start with low-barrier charts. Many small-language markets have dating or social Top Grossing charts with only a few dozen spots. A new app can buy a small amount of traffic, collect a few payments, and chart within days. Charting is itself exposure, and the cheapest way to validate a payment model. Optimize ads on payment events, and use visible novelties like game rooms and maps as hooks for creatives. Don't use "verified real people" claims that contradict the terms.

When to act, and when to wait: no rush, and there shouldn't be. Spend two or three weeks watching the target market's dating Top Grossing chart to see how many new apps enter. In Turkey, this window is already filled by this wave.

What would show this call was wrong: ① Aramızda drops off Turkey's dating Top Grossing chart within a month, with installs stalling around two thousand. That would mean even this latecomer hasn't paid back its ad spend, and the opportunity is smaller than it looks. ② Conversely, if it and similar apps like Fubby climb to tens of thousands of installs and reach the top 20 within a month or two, Turkey's coin-chat market isn't saturated yet, and this call was too conservative. ③ If Google Play pulls apps like this in bulk from the Turkish store, or explicitly bans undisclosed AI profiles in its dating policy, the path is closed by the platform, and the reskin business is over.

Android store page ↗

Research notes: competitor table, acquisition evidence, full research report
Competitors
NameWhereLaunchedScaleHow it differs from this one
Civarda Sevgili BulAndroid2021-112.3M installs, Turkey Dating Top Grossing #4 (googleplay library 11-18)Also targets people nearby, an established coin-chat app with lots of fake-profile complaints; Aramızda adds neighborhood-level location, a map, and game rooms
Kaçamak Sevgili BulAndroid2022-091.61M installs, #9 on Top GrossingEstablished peer, no game rooms
Yakınlarda: Konumdan SohbetAndroid2025-0878k installs, Top Grossing #55Also shows people nearby on a map, the closest concept; no six-player game rooms
FubbyAndroid2026-08-2549k installs, Top Grossing #21; 229 TikTok adsA Turkish team using a US LLC as a shell for a coin-chat app, with far larger ad spend than Aramızda
Datte / Hovarda(Capraz Innovations)Android2026-08 / 2026-056.5K / 23k installs, Top Grossing #48 / #22One person running two reskinned brands, Meta ad buying optimized for payment value, pure coin chat
Alev / Çapkın(Volyra LLC)Android2026-04 / 2026-0835k / 12k installs, Top Grossing #35 / #44Two brands under one LLC, with simpler mechanics
Spin the Bottle(Cyboma)Android2016-1113.6M installsThe prototype for multiplayer bottle-spin rooms, positioned as a social game; Aramızda puts it inside a dating app
WePlayOtherAround 20192024-06 Turkey app revenue, #10 (ensonhaber)Voice game-room socializing, with large real-person scale
Tinder / Badoo / BumbleBig-company features2010–2015Tinder 530M and Badoo 410M installs, Turkey Dating Top Grossing #1 / #7 / #3Mainstream real-person platforms; Aramızda is the local small-language long tail
Where users come from
  • Inferred Paid acquisition (Meta / Google, with conversion events sent back): the privacy policy lists Meta (Facebook) / Google Analytics for measuring ad conversions; Google's Ads Transparency Center has verified US advertiser APPION DIGITAL LLC with about 3 ads, creatives hidden for age restriction; Meta Ad Library search by name / package returns nothing
  • Inferred Store search (Turkish-language keywords in the title): the title "Aramızda: Flört & Arkadaşlık" uses the two big terms flört and arkadaşlık
  • Verified: multi-country store spread / low-bar grossing charts: Aramızda sits at #49 on Turkey's dating Top Grossing chart, which has only 74 apps; it also ranks #153 on Nigeria's dating Top Grossing; in a recall, it also showed up on AppBrain's new dating charts in Austria and Denmark (unverified)
  • Inferred: content / creators / media: x-radar, Jina and Brave found no related coverage

Aramızda: Flört & Arkadaşlık (com.sparkin.app): why it charted

Verdict

Strictly speaking, it hasn't taken off yet. Launched 2026-10-16, and as of 11-18 it has 1,827 installs worldwide. "#49 on the Turkey Play dating Top Grossing chart" sounds good, but that chart has only 74 apps in 11-18, so ranking #49 just means "some paying users."

It's one of the 2026 wave of Turkish reskins built on the same formula: a US LLC shell + Turkish-language + coin/membership chat dating. It makes money from three things working together:

  • a paywall (Gold membership + coins);
  • "AI-managed virtual profiles" disclosed in the terms of use, which chat with male users;
  • paid acquisition with paid-event feedback. The privacy policy says data is sent to Meta for ad conversion; the developer has a verified Google advertiser account with about 3 ads.

The store-page differentiators are neighborhood-level local discovery, a map mode, and six-person real-time game rooms like bottle spin, hot seat and blind match. But with just over a thousand installs, real people can't fill six-person rooms, so these features aren't why it charted.

Fundamentals (data)

  • Developer: Appion Digital LLc. The database has only this one app; the website is sparkin.online, support email [email protected] (googleplay database app table).
    • Google Ads Transparency Center finds "APPION DIGITAL LLC": verified identity, based in the US, with an estimated ~3 ads. The page notes "some ads contain age-restricted content and require login to view," so creatives aren't visible (adstransparency.google.com/advertiser/AR09039705534332141569, Scrape Creators advertiser search results).
  • Installs (googleplay database app_snap): 1,027 (11-16) → 1,027 (11-17) → 1,827 (11-18). AppstoreSpy estimates about 93 installs/day from 11-13 to 11-15, 11-15 lifetime 1,027.
    • Averaged over its first month, that's about 30 installs a day, with a single-day spike of +800 on 11-18. It could be paid acquisition ramping up, or Play batch-updating install counts; the two can't be told apart.
  • Charts (googleplay database chart_rank):
    • Turkey dating Top Grossing: 55 (11-16) → 52 (11-17) → 49 (11-18);
    • Nigeria dating Top Grossing: 171 → 163 → 153;
    • It has never charted on any Top Free chart. In other words, few installs, but people are paying.
  • Chart bar: 11-18 Turkey's dating Top Grossing has only 74 spots. On the same day, #23 Gönlünegöre had only 630 installs, #48 Datte 6,511 installs, #61 Denk 1,497 installs. The bar is very low.
  • Ratings / reviews: None (0 entries in the database).
  • Store page (Play, hl=tr): PEGI 18, contains in-app purchases. Gold membership features: see who liked you, unlimited likes, Boost, invisible browsing, voice and video calls.
  • Store page selling points: "Blue-check selfie verification makes fake profiles disappear" and "%100 güvendesin" (100% safe).

Its "secret": the terms say there are AI virtual profiles

  • Terms of use, section 3 (sparkin.online/kullanim-sozlesmesi.html, updated 2026-08-11, two months before the App launched, suggesting a ready-made template):
    • The platform has "AI-managed virtual profiles" that do not represent real, verified people, and AI models automatically reply in chats;
    • Continuing to use the app means accepting that "there may be virtual profiles on the platform, and they are not real users."
  • Privacy policy (sparkin.online/gizlilik-politikasi.html, 2026-10-17):
    • Chats with virtual profiles are forwarded to large models via OpenRouter;
    • Other stack: Firebase, Cloudflare R2, Agora (audio/video), Google Play billing;
    • Explicitly lists Meta (Facebook) / Google Analytics, used to "measure ad conversions".
  • The website homepage also says "the platform may include AI-supported virtual profiles."
  • This is the opposite of the store-page promise of "goodbye fake profiles, 100% safe": the store page speaks to users, while the terms speak to reviewers and legal disclaimers.
  • External context: Anthropic's threat report 2026-09 disclosed that a Chinese developer used 20+ AI personas across 4,700 dating apps to chat with users and sell coins, with about 75% of profiles being AI. Apple and Google have been notified (Turkish-language report: wired.com.tr, 2026-09-16).
    • Aramızda writing AI profiles into its terms looks like a hedge against this kind of scrutiny. But the store-page marketing contradicts the terms, so removal risk remains.

Competitors (doing the same thing)

Turkey is a high-ground for paid social and dating apps. 2024-06 Turkish apps in the top ten by revenue include Azar, Tinder, Tango, LivU and WePlay, and in the month in question these apps together grossed over $3M (ensonhaber.com, 2024-07-03).

Turkey's dating Top Grossing chart has a whole crowd of "Turkish-language coin / membership chat" apps like Aramızda. The table below uses scale figures from the googleplay database 11-18:

ProductPlatformLaunchedScaleHow it differs from Aramızda
Civarda Sevgili BulAndroid2021-112.3M installs, #4 on Turkey dating Top GrossingAlso pitches "people nearby" with old-school coin chat; Şikayetvar is full of fake-profile and paid-complaint reports
Kaçamak Sevgili BulAndroid2022-091.61M installs, #9 on Top GrossingVeteran in the same category
Yakınlarda: Konumdan SohbetAndroid2025-0878k installs, Top Grossing #55Also "see people nearby on a map"; closest concept
FubbyAndroid2026-08-2549k installs, Top Grossing #21Turkish team + US Missouri LLC; 229 TikTok ads (stringers product database)
Datte / Hovarda(Capraz Innovations)Android2026-08 / 2026-056.5K / 23k installs, Top Grossing #48 / #22A one-person double-branded reskin; Meta paid acquisition optimized for paying users
Alev / Çapkın(Volyra LLC)Android2026-04 / 2026-0835k / 12k installs, Top Grossing #35 / #44Two brands under the same LLC
Pawn, Tenhada, KiDate, Bontik, Gönlünegöre, Denk, etc.Android2025-09 to 2026-10Hundreds to tens of thousands of installs, all on the 74-spot grossing chartReskins from the same 2026 wave
Spin the Bottle(Cyboma)Android201613.6M installsThe prototype for multiplayer bottle-spin rooms; Aramızda packs this mechanic into a dating app
WePlayAndroid / iOSAround 2019#10 among 2024-06 Turkish apps by revenueVoice game-room social
Tinder / Badoo / Bumblebig tech2010–2015Tinder 530M and Badoo 410M installs, Top Grossing #1 / #7 / #3Real-person scale, the mainstream choice
Azarbig tech—#2024-06 in Turkey's dating revenueRandom video chat

Followers: I haven't seen anyone copying Aramızda yet. It is itself a latecomer to this wave; a dozen or so nearly identical products launched in the same period.

Where users come from (growth levers)

  1. Paid acquisition (likely, but creatives not visible):
    • The privacy policy explicitly says Meta / Google Analytics are used to "measure ad conversions";
    • The developer has a verified Google advertiser account with about 3 ads, creatives hidden by age restriction;
    • The install curve shows no signs of organic word of mouth.
    • Similar apps Datte and Fubby have been confirmed as Meta / TikTok paid acquisition optimized for paid events (stringers product database).
    • Searches of the Meta Ad Library for "Aramızda Flört", "com.sparkin.app" and "Semtindeki insanlar" found nothing; our metaads / tiktokads / googleads databases don't track it (it isn't on the list).
  2. App Store search: the Turkish-store title "Aramızda: Flört & Arkadaşlık" uses the two big search terms "flört" and "arkadaşlık".
  3. Content / creators / media: nothing found. x-radar: 30 days, 0 items. Jina / Brave searches for "Aramızda" only turned up a same-name iOS app, "AI İlişki Koçu" (a different developer), and unrelated content. TikTok was not checked (the TikHub legacy account ran out of balance).
  4. Multi-country spread: it showed up during recall on AppBrain's Austria and Denmark "new free dating app" charts (unverified: the AppBrain pages have a human-verification check). It also appears on the Nigeria dating Top Grossing chart, most likely a long-tail app that squeezes on with a small amount of paid installs.

Where the money comes from

  • Once male users are in, "who liked you", "voice and video calls" and "send gifts" all require Gold membership or coins.
  • The Turkish dating market has more men than women. AI-generated virtual profiles make male users feel "someone replies", and the paywall collects the money.
  • Paid acquisition cost is recovered from the first purchase, with optimization based on paid events sent back to the ad platform.
  • Over a thousand installs was enough to reach Top Grossing, which suggests a decent payment rate. But this chart has only 74 slots and a low bar, so absolute revenue is very likely small.

Differentiation

The store-page differentiation has three parts:

  • Local matching down to the "district / neighborhood (il / ilçe / semt)" level;
  • Map mode plus stealth mode;
  • Real-time six-player game rooms (spin the bottle, hot seat, blind match), plus two-player mini-games in chat (this or that, compatibility test, truth or dare).

These selling points are not why it charted:

  • With a thousand-plus installs, real people can't fill six-player rooms, so AI profiles most likely fill them in.
  • It charted on the same combination as its peers: paywall + AI chat companion + paid acquisition.
  • The game rooms work more as new hooks for ad creatives and the store page.

Takeaways

  • Find low-bar Top Grossing charts: a small-language dating Top Grossing chart has only a few dozen slots, so a new app can chart within days. Charting brings exposure, and it's a cheap way to validate a payment model.
  • Template-based fast reskinning: the same stack (Firebase + R2 + Agora + OpenRouter) and the same agreement template can open a new market just by swapping the name and language. A US LLC serves as the developer entity, verified by Google as an advertiser.
  • Compliance hedge: writing AI virtual profiles into the user agreement is meant to deal with platform scrutiny of "AI fake-human companions". But the store-page marketing contradicts the agreement, so this step is itself a risk point.
  • What others should not copy: taking money by having AI pass as real people will draw close platform attention after Anthropic's 2026-09 report, and carries consumer complaint and legal risk.

Unknowns

  • Actual revenue and payment rate: the Top Grossing chart gives no amounts, only that revenue exists.
  • Ad creatives, targeted countries, and money spent: Google creatives are hidden due to age restrictions, and Meta search returns nothing.
  • Whether the +800 installs on 11-18 came from added spend or from a Play statistics update.
  • Who operates it: recall says sparkin.online belongs to Sparkin Online Service LLP in Tamil Nadu, India. But ScamAdviser shows the domain was "registered only recently", and that LLP was registered 2024 years ago with the business scope "other educational services". It looks like a name coincidence, and there is no evidence linking the two. The site, agreement and KVKK compliance are all in Turkish and governed by Turkish law, which looks more like a Turkish team using a US LLC as a shell (the same approach as Fubby).
  • Whether the six-player game rooms have real people in them, and what share of them are AI profiles.

Sources

This article was written by AI from public data and the data we collect daily. Numbers are sourced, and inferences are labeled.