AI Girlfriend 18+ Apps: Age Checks and Content Rules

There's an app on Google Play called "AI Girlfriend 18+". The US store rates it Mature 17+.

That's the shape of this whole search, so here's the direct answer. Every AI girlfriend 18+ app ranking for this query says it's for adults. On five of the six, the thing standing between a visitor and the product is a button, a checkbox or a sentence in the terms saying you affirmed your age by showing up. On one, LoveForever, there's a third-party age check with a named provider, and it applies only to visitors from the UK, the EU and other places that have passed a law requiring it. Everyone else gets the checkbox. Once you're in, the content rules range from a detailed list with a named scanning step to a single sentence about trafficking, and two of the six describe no moderation method at all.

Everything below was read from each app's own terms, policies, page source and store listing on 15 September 2026. Prices aren't the subject here, and the privacy side, meaning what the apps do with the conversation afterwards, is a separate question covered in the adult companion privacy comparison. This page is about the door and the house rules.

What Backs the Number

Six apps rank for this search. Seven of the eight organic results are the vendors themselves or a developer promoting their own product, so the only "18+" on the results page is the sellers' assertion. Here is what each assertion rests on.

Read from each vendor's own documents on 2026-09-15. Document dates are the ones printed on the documents.

The column that matters is the fourth one. One app in six has a mechanism that could, in principle, tell an adult from a 16-year-old, and it applies to a subset of visitors. The other five have a record that somebody clicked. That's not a criticism unique to these six. The mainstream platforms mostly do the same, and the exception there, one operator publishing outcome numbers from a Yoti-backed programme, is in the sexting policy audit. What's different about this shelf is that "18+" is in the product names and the URL slugs, so the gap between the label and the mechanism is the product's own marketing.

HeraHaven's Gate Is Four Lines of JavaScript

HeraHaven ranks first, and its page source is the most instructive document in this article, because it shows exactly what an age gate on a companion site is.

Load the homepage and an overlay covers it. "User Verification," it says, then "By entering, you confirm you agree to our Terms of Service, Privacy Policy, and are at least 18 years old or the age of majority in your jurisdiction," then a button labelled Enter. The button's code removes the overlay and writes a flag to the browser's local storage so you aren't asked again. That's the check.

A few lines further down there's a script that decides whether to show the overlay at all. It skips it if the page address carries a particular query parameter, and it skips it if the browser identifies itself as Googlebot, bingbot or DuckDuckBot. The first exception is presumably for the company's own testing. The second means search engines are shown the site without any gate, which is standard practice for crawlers and also a plain statement of what the gate is for. It's for the human, as a liability record, and not for the machine that indexes the content.

The terms behind the button are more interesting than the button. HeraHaven's terms page is a shell that redirects to one of two documents depending on where you are, one for the European Economic Area and one for everywhere else, both marked "version: 22nd of August, 2026", both naming a company at an address in Nicosia, Cyprus. Both say you must be 18. Both say your content may be used to train the company's models unless you opt out by email. Both say every charge is non-refundable except where the law requires otherwise. And the app's code contains a page called verification-status that waits for an age verification flag to flip. So some verification flow exists, for someone. Nothing on the public site says who triggers it, what provider runs it, or whether it's a regional requirement like LoveForever's. I'd treat that as a question for their support address rather than a feature.

LoveForever Verifies Some Visitors and Says So

LoveForever's terms, last updated 22 April 2026 and labelled version 3.1, are the only document in this set that describes an age check with a provider, a method list and a data flow.

Section 2.1 is titled for UK and EU users. Visitors from the United Kingdom, the European Union and "other regulated jurisdictions" have to verify through Go.cam, by one of four methods, a selfie scan using AI age estimation, a government ID scan, a credit card check, or email verification. The site says it receives only a pass or fail, never the document or the face, and that the verified status is kept for a year. Then the sentence that sorts the world in two. "Users in non-regulated regions may use simple age confirmation, though we reserve the right to require full verification for any user at any time."

So an adult in Manchester scans a selfie and an adult in Ohio ticks a box, for the same product, under the same terms. That's the honest state of age assurance on the web in 2026. Where a law exists, a check exists. Where it doesn't, a checkbox does. LoveForever's footer puts it more grandly, describing Go.cam as a certified partner and adding "No minors. Always verified." The terms say always is a regional word.

The same document is specific about what's forbidden once you're in, and specific about how it looks. Section 4.3 says "All prompts are scanned before processing" and lists zero-tolerance categories, anything involving minors, non-consensual content, violence and weapons, extremism and hate speech, kidnapping and trafficking, gore, self-harm, and illegal drugs. Violations get a warning, then suspension, then termination, with an email address for appeals. Voice calls get their own section, which says the audio is recorded by a voice partner, VAPI, transcribed by Deepgram and summarised by OpenAI, with key facts kept as memories you can ask to delete. I'd rather know that than not, and most of this category doesn't tell you.

And then section 14, which says the terms are governed by "the laws of [Your Jurisdiction]" and disputes go to "[Arbitration Association]". Square brackets included. Somebody built this document from a template and never filled in the two blanks that decide where you'd sue. It doesn't change the age check. It does tell you how carefully the rest was proofread.

The App Called 18+ Is Rated 17+

The fourth result is a Google Play listing, and the app's name is the search query. "AI Girlfriend 18+", published by Amacci Technologies of São Paulo, with 100,000-plus downloads, ads, in-app purchases, and an update on 24 April 2026.

Open it from the United States and the content rating is Mature 17+, with the descriptor Sexual Themes. Open it from Europe, as I did, and the same listing shows PEGI 16 with the descriptor Sex. Google's own developer documentation says ratings are "assigned by separate rating authorities and determined by your questionnaire responses", so this isn't Google contradicting the developer. It's the developer's own answers producing a rating one year below the number in the app's name on one continent and two years below it on another, on a scale that does have an Adults Only tier for 18 and up. The listing's description says the app is "intended exclusively for users 18 years and older" and, a line earlier, that "Both IntimAI and the Play Store are not responsible for actions taken within the app". The three layers that are supposed to keep a store app adults-only, and why none of them is a check, are laid out in the age gate audit. This listing is a clean example of layer one.

One more thing on the same page. The data safety panel declares no data collected, no data shared with third parties, and "Data can't be deleted". The privacy policy it links to is a Blogspot post from October 2024, written in the first person as "I, IntimAI", which lists name, age, email address, device identifiers and payment data as collected, and "Display personalized ads" as a use. A panel that says nothing is collected, beside a policy that lists what is, beside a listing that says the app contains ads. I'm leaving the privacy comparison to the sibling article, but that contradiction sits on the store page of the app named for this search, and it belongs here.

Three Affirmations, Three Adjectives

Kupid, Luvr and Romantic AI all run the same mechanism, a statement that you're 18, and describe it with escalating confidence.

Kupid's terms carry a revision date inside square brackets, "[ 10/08/2026 ]", and section 2.2 says users "affirm that they meet the minimum age requirement" by using the platform, followed by the sentence that the company "cannot be held responsible for any inaccuracies or misrepresentations regarding user age". That's the same disclaimer Candy.ai's underage policy uses, almost word for word, and I'd read it the same way, as an accurate description of what an affirmation does. Payments run through a Cyprus company, the terms are governed by Cyprus law, and every sale is final. The full product is in the Kupid review.

Luvr publishes an Underage Policy, effective 1 January 2024, that says the company uses "age verification mechanisms, such as an initial age gate", and that each new user "must confirm that they meet the age requirement". The terms, dated 15 January 2024, say "You represent that you are over the age of 18" and define the governing law as that of "the Country", which the definitions section says "refers to: San Francisco, United States". The homepage HTML served to my request contained no age prompt at all. If a gate appears, it's drawn by script after the page loads, and I can only report what the server sent.

Romantic AI has the most documents and the most confident adjective. Its Content Management Policy, dated 26 May 2026, describes an "Enhanced Age Gate" and "an advanced age verification gate" designed "not merely as a barrier but as a proactive filter". The next paragraph defines it. "Users are required to affirmatively declare their age." The Underage Policy of the same date calls it "an initial landing page or age gate" where "Users are prompted to affirm that they are above 18". And the terms, updated 10 July 2026, say what the other two documents don't. "We cannot guarantee that each user is at least the required minimum age." I'd give the company credit for that sentence. It's the true one.

What You Can't Do Once You're In

The second half of the search is the content, and here the six apps separate more than they did at the door.

Compiled from each app's terms and content policies, read 2026-09-15.

Two of the six, LoveForever and Romantic AI, publish something a reader could hold them to. LoveForever's list is shorter and its method is more concrete, since "scanned before processing" describes a point in the pipeline. Romantic AI's list is longer and includes the two categories I'd look for first on any adult platform, content resembling minors and impersonation of real people, and its policy is candid that monitoring is not continuous and that reported content gets assessed within five business days rather than instantly.

Luvr's list exists and is dated January 2023, which makes it the oldest document in this article and the one I'd most expect to have drifted from the product. Kupid describes a method without a list. HeraHaven and IntimAI describe neither, and HeraHaven's terms put the responsibility on the user in the general way most terms do, without naming what would get an account closed.

Questions

Does 18+ mean the app checked? On these six, no, with one regional exception. LoveForever runs a Go.cam check for UK, EU and other regulated visitors. Everything else is an affirmation, and HeraHaven's page source shows what an affirmation is, a button that removes an overlay.

Is an 18+ app on Google Play rated 18+? The one named for it isn't. Mature 17+ in the US and PEGI 16 in Europe, from the developer's own questionnaire answers.

Do UK and EU users face different checks? On LoveForever, yes, by its own terms. The other five don't split their rules by region on the pages I read, though HeraHaven publishes separate EEA and rest-of-world terms with the same age clause in each.

What gets you banned? On LoveForever, any of eight listed categories, with warnings first. On Romantic AI, any of eight, assessed within five business days. On Luvr, five. On the others, whatever the company decides, because there's no list.

Can the app tell if a character looks underage? Romantic AI and LoveForever say they prohibit it and act on attempts. Luvr prohibits exploitative content toward minors. The other three don't address it in the documents I could load, which is the gap I'd weigh most heavily.

What about the conversation itself, afterwards? That's the privacy question, and it's a different set of documents. The sibling comparison of adult companion apps on data, retention and model providers is at the adult companion privacy comparison.

What I'd Do

If the age check matters to you, and for a shared device or a household with teenagers it should, LoveForever is the only one of these six that describes a real one, and only if you're in a region that forces it. Elsewhere it's a checkbox on every app on this list, and the honest move is to stop reading the "18+" in the name as information.

If the content rules matter, read the blocked content policy before the terms, because the terms on this shelf are mostly boilerplate and the policies are where the categories live. Romantic AI's and LoveForever's are the two worth ten minutes. If an app has no such document, as HeraHaven and IntimAI don't, assume the rules are whatever the moderation model decides on the day.

And read the document dates. A content policy from January 2023, a terms page with the jurisdiction left as a placeholder, a revision date in square brackets. None of those is a safety problem on its own. Together they tell you how much attention the paperwork gets, and on a product that exists to be adults-only, the paperwork is the product's only proof.