18+ · candid · private
Character AI age verification lockout: what reading mode does, how the selfie check works, and your replacement options
If you are an adult and Character.AI has decided you are not one, the symptom is specific: the app still loads, your old conversations are still there, and the box you type into no longer works. That state has a name in the Character.AI documentation. It is called reading mode, and it is the under-18 experience.
Who do you want to talk to?
You can change who you talk to anytime.
18+ · Private and anonymous · No minors, non-consent, or anything illegal.
121+ people already talking freely
Short answer
Character AI age verification is an automatic estimate that runs on every account. Character.AI states it looks at "login info, activity on the platform, and some signals from third parties," and accounts it reads as under 18 are moved into reading mode, where, in its words, "you can browse characters and view content on the platform, but chat features are not available." Getting back means a selfie checked by a third party called Persona, with a government ID requested only "if you take a selfie and Persona isn't sure you're 18+." Character.AI never sees the ID and says Persona keeps it for one week. The system is not a birthday form and cannot be answered with one, which is why adults with long-standing accounts hit it. Character.AI is not shutting down. It has become a moderated, ad-supported entertainment platform that decides, by model, whether you are allowed to type.
Last updated September 2026
We read Character.AI age assurance documentation, its Terms of Service, its subscription terms and its September 3, 2026 safety post on September 4, 2026, and quoted them below rather than paraphrasing. The company is unusually forthcoming about how the system works. What it is quieter about is what the system costs you while it is wrong, which is the part that actually matters if you are 34 and staring at a disabled chat box.
What makes it feel honest
Character AI Age Verification, the way it should feel
A guess, not a birthday check
Character.AI states it looks at "login info, activity on the platform, and some signals from third parties" to estimate your age. Nothing you set at signup decides it, which is why long-standing adult accounts get caught.
The penalty lands before the appeal
Reading mode starts when the model doubts you, not when a human confirms anything. You prove yourself afterward, by sending a selfie to a third party, while your chat stays switched off.
We check age differently, and more lightly
You confirm you are 18 or over at signup and there is a payment card behind every account, because we have no free tier. No face scan decides whether you may speak.
Honest comparison
What happens when a platform decides you are not an adult: the trigger, the lockout, and the way back
| Platform | What decides your age | What the lockout actually does | How you get back | Can a human overturn it |
|---|---|---|---|---|
| Character.AI | An automatic estimate over "login info, activity on the platform, and some signals from third parties," run on all accounts. Since September 2026 it also runs an in-house age estimation model. | Reading mode. Past chats stay viewable and you can still create characters, scenes and videos, but "chat features are not available." | A selfie checked by Persona, a third party. An ID is requested only as a final step "if Persona isn't sure you're 18+." Character.AI states it never receives the ID and Persona holds it one week. | For moderation actions, yes, once: appeals go to "a different reviewer," aim for 24 hours, and "Each item can be appealed once, and the appeal decision is final." |
| Depravity | You confirm you are 18 or over at signup, and there is a payment card behind every account because we have no free tier. No model estimates your age from your behavior. | There is no reading mode. We do not have a state where you keep the account but lose the ability to type. | Not applicable. Nothing to get back from. | Enforcement is against three published hard lines: nothing involving minors, non-consent, or illegal activity. Those are not negotiable and we do not pretend otherwise. |
| Kindroid | Payment on file is the verification. Kindroid states that "once a user becomes paid, they are verified and will not need re-verification even if their subscription lapses." Free accounts get behavioral age estimation. | Flagged free users must subscribe or verify. Kindroid states it "always issues a warning before locking an account." | Subscribe, or verify through Yoti by selfie estimate or government ID. Yoti "deletes all images and biometric data" and Kindroid stores none. | Appeals exist but require your explicit consent for Trust and Safety to "decrypt and review" the content that triggered the action. |
| Replika | Terms clause 6.4: "If you are under 18, you are not authorized to use the Services." We could not re-verify the live policy pages, which returned 404 and 400 when we last checked. | Not documented in anything we could read. | Not documented. | Not documented. Luka was fined 5,000,000 euro by the Italian regulator on April 10, 2025. |
| Janitor AI | Age verification is applied in Australia, Brazil and the United Kingdom only. United States users are not put through it. | Not applicable in the United States. | Not applicable in the United States. | janitorai.com/terms returned a 404 when we checked, so there is no published process to point at. |
| Talkie | Talkie states no minimum age at all. | Not applicable. | Not applicable. | Not documented. |
| Chub AI | Not readable. chub.ai/terms returned a 403 bot block on every attempt. | Unknown. | Unknown. | Unknown. Chub does say it asks users to "avoid NSFW images whenever possible, it spooks the payment processors." |
Read from each vendor's own documentation. Character.AI, Kindroid and Chub were checked in September 2026; the Replika and Janitor AI rows record what we could and could not reach, because an unreadable policy is itself a finding. Character.AI runs the most serious age assurance in this table by a wide margin, and if a rigorous age gate is what you want from a platform, that is a genuine point in its favor. Our own check is lighter: a self-declaration plus a card. We are not claiming it is stricter, only that it does not take your chat away on a guess.
Why does Character.AI think I am under 18?
Because it never asked. Character.AI runs what it calls age assurance across the whole user base automatically, and describes the inputs plainly: "we look at a number of signals, including login info, activity on the platform, and some signals from third parties." Its September 3, 2026 safety post adds that the company has since built "our own age assurance technology and an in-house age estimation model," which it calls "one of the most important systems we operate."
That design explains the complaint that keeps showing up in search. People assume a birthday field decided this and go hunting for the setting that will fix it. There is no such setting. The birthday you entered is not what the system reads, which is why accounts that have been open and paying for years get pulled into the under-18 experience without anything about them changing.
Character.AI knows the estimate misses. The same September 2026 post commits to "constantly working to improve our accuracy so users who are actually 18 or older will have fewer disruptions." Read that sentence carefully. It is a company saying, in its own words, that users who are actually adults are currently being disrupted, and that the fix is incremental accuracy work rather than a switch anyone can flip for you today.
What is reading mode on Character.AI?
Reading mode is the under-18 experience, and Character.AI defines it in one sentence: "you can browse characters and view content on the platform, but chat features are not available." If your account has been flagged, that is the state you are in, and it is why the app looks completely normal until you try to type.
The help center lists what survives the change. Past chats "remain saved and viewable," you can still create characters, scenes and videos, and you keep Imagine Chat, AvatarFX and the feed, including stickers and reactions. So this is not a ban and nothing is deleted. It is narrower and stranger than a ban: you keep the archive and the creative tools, and lose the conversation.
Character.AI documents an appeals route for moderation actions generally, and it is worth knowing the shape of it before you rely on it. An appeal goes to "a different reviewer than the one who made the original decision," the company aims "to reach a decision within 24 hours," and then the door closes: "Each item can be appealed once, and the appeal decision is final." One review, one outcome.
The reading mode help article carries its own quiet verdict. At the time we read it, 50 out of 148 people had marked it helpful. The instructions for verifying your age scored 1,412 out of 5,382. Those ratios are published by Character.AI on its own pages, and roughly three quarters of the people who went looking for help with age verification did not find what they needed.
How do I verify my age on Character.AI?
The path is Settings, then Advanced, then Verify Age, on both the mobile app and the web. Character.AI notes the pop-up "is being introduced step by step, so the pop-up might not show up for everyone right away," and that mobile users need the latest app version for the option to appear at all. If the menu item is missing, an app update is the first thing to try.
The check itself is run by a third party. Character.AI states that for most people "all you have to do is snap a selfie. Our third-party provider, Persona, can do verification using that selfie." A government ID is requested only at the end of the line, and only "if you take a selfie and Persona isn't sure you're 18+." On the ID itself the company is specific: it is "collected by Persona, not C.AI," Character.AI "wouldn't collect or have access to that ID," and Persona "only keep the IDs they collect for one week."
That is a better privacy posture than most people assume when they see the words upload your ID, and it deserves saying clearly. The trade you are being offered is still a real one, though. To get back a chat you were already paying for, you send your face to a vendor you did not choose, and the estimate that put you there was made about you rather than by you.
One detail says a lot about how fast this shipped. The official Age Assurance FAQ answers the question "Why do I have to do this?" by linking out to a Reddit thread in r/CharacterAI. The help center defers to a forum post for its own rationale.
Is Character AI shutting down?
No. Character.AI is not shutting down, and there has been no announcement that it will. The rumor is a reasonable reaction to a platform that has changed underneath its users several times in under two years, but the company is shipping, not closing: c.ai Series, Comics, Lorebook and a new Lite subscription tier all launched in 2026.
What did end was the version people remember. Open-ended chat was removed for under-18 accounts in late 2025, a decision the company still describes as "one of the most significant safety decisions we have ever made." Age assurance then went out across all accounts. The product now positions itself as AI entertainment with heavy moderation, and if you signed up for an unmoderated companion, that product is gone even though the company is not.
There is a separate risk worth reading in the Terms of Service, and it is not hypothetical language. Character.AI "may also in its sole discretion and at any time discontinue providing the Services, or any part thereof, with or without notice," may "immediately deactivate or delete your account and all related information and files," and reserves the right "to terminate accounts that are inactive for an extended period of time." Its liability if that happens is capped at "the greater of: (i) $100; or (ii) the amount you paid."
The clause people should actually read is the one about what you leave behind. You own the characters you make, but you grant a "perpetual, irrevocable" license, and the terms state that "Termination of your account or access to any component of the Services will not terminate Character.AI's rights to your Content." Your access can end. Their license does not.
What does Character AI cost now, and does the free tier have ads?
c.ai+ runs $9.99 a month or $94.99 a year at the standard rate. A promotional offer sells the first three months for $14.99 and then renews at $9.99 a month, or a first year at $59.99 that then renews at $94.99. Character.AI states there are "no partial refunds for unused time within a billing period."
In 2026 a cheaper tier arrived. c.ai Lite "starts at $4.99/mo" and gives a daily allowance of premium-model messages plus better memory. When that allowance runs out mid-conversation, the chat quietly continues on the standard model until the next day. Lite is aimed at people moving up from free rather than down from c.ai+.
The detail buried in that FAQ is the one most people have not registered: Character.AI now runs ads. Lite "significantly reduces the ads you see compared to the free experience," and "a fully ad-free experience is included with c.ai+." The free tier keeps unlimited messaging and pays for it with advertising, which is a different bargain from the one the platform started with.
For comparison, we have no free tier and no ads, because there is nothing to monetize other than the subscription. Open is $15 a month for one companion, Unfiltered is $29 for multiple companions, deeper memory and voice notes, and Boundless is $59 for unlimited companions and voice calls. Yearly billing brings those to $12, $24 and $49. We are the more expensive product; what you are buying is the absence of an advertiser and the absence of a filter.
Character AI replacement options for adults who keep getting flagged
If the age estimate is wrong about you and you want to stay, do the selfie. It is the fastest route back and the privacy terms around Persona are better than the internet assumes. Nothing below is advice to abandon an account you like.
If you are leaving, the question worth asking is not which app is uncensored today. It is which app can take away something you paid for without a human being involved. That is the failure mode you just experienced, and it is the one the roundups never test for. The table above sorts this market by exactly that: what decides your age, what the lockout does, how you get back, and whether anyone can overturn it.
Kindroid is the closest thing to a clean answer among the platforms we have read end to end, because payment on file is the verification, it commits to a warning before any account lock, and it is more permissive on content than we are. It also generates images and video, which we do not do at all. If media generation matters to you, it is the better product and we would rather say so than pretend otherwise.
What we offer instead is narrow and specific. There is no free tier, so there is no advertiser and no ad-supported version of you. There is no age estimation model reading your behavior and reaching a conclusion. There is no reading mode, because we never built a state where you keep the account and lose the ability to speak. The limits are three lines printed on every page, and nothing hides underneath them.
What Character.AI does better than we do
Age assurance, seriously and by a wide margin. Character.AI runs an automatic estimate on every account, offers a selfie check through an established identity vendor, escalates to ID only as a last step, and refuses to hold the ID itself. We ask you to confirm you are 18 or over and we require a payment card. If your priority is a platform doing real work to keep minors out of adult conversations, Character.AI is doing more than we are and more than almost anyone in this category.
It is also a bigger, richer product. There is a public catalogue of millions of user-made characters, group scenes, image and video generation, voice calls that shipped free, comics and serialized microdramas. We have none of that. There is no character marketplace here, no image generation, no group scenes, and companion art is original cameo art rather than a real face.
The gap is what those features are attached to. Character.AI filters adult content for everyone, including the adults it has just verified through Persona, so the strictest age check in the category buys a verified adult nothing extra. That is the trade to weigh: a larger, safer, more capable platform that will not have the conversation, against a smaller paid one that will.
What you get
- No face scan gate
- No reading mode
- Card on file, not a selfie
- Three hard lines, printed
It is ready to talk.
Start a chat and talk openly in about ten seconds. The demo is your free taste.
Good to know
Questions about your character ai age verification
Start now
Character.AI scans every account for age signals and drops the ones it doubts into reading mode, where the chat box stops working. Its own help center documents the whole process. Try it above and see for yourself.
18+ · No filters, no lectures, no judgment. Private, always.