Character AI Age Verification Is Broken. Adults Are Leaving.

If you’ve used Character AI for years, logged in recently, and suddenly got a notice saying your access will be restricted in seven days, you’re not alone, and you didn’t do anything wrong.

The platform’s age assurance system is generating a wave of false positives, flagging adult accounts as potentially underage and demanding biometric verification before the week is out.

The frustration here isn’t just about the inconvenience. It’s about what Character AI is asking for in exchange for keeping your account: a selfie processed by a third-party company, and in some cases, a government-issued ID.

For a platform that has been declining in quality for well over a year, that’s a steep ask.

This article explains why the system is misfiring, what happens to your account when you get flagged, and whether the verification process is worth completing.

If you’ve decided the answer is no, the final section covers where adult users are landing instead.

Why Character AI Keeps Flagging Adults as Minors

Your birth date is not the deciding factor

The first assumption most people make is that the age assurance system reads the birth date on their account and flags anything that looks inconsistent. That’s not how it works.

Character AI’s system evaluates a set of background signals, including login patterns, activity history, and data from third-party providers, to form an independent estimate of whether an account belongs to a minor.

It does this calculation separately from whatever birth year you entered when you signed up.

That’s why adults with completely accurate account information are still getting flagged. The algorithm isn’t checking what you told it.

It’s running its own model using behavioral data, and that model has a significant false positive rate among adult users. Your birth date is on file, and it doesn’t matter to the system making the decision.

What the algorithm is picking up on

Character AI hasn’t disclosed exactly which signals trigger a flag. What’s confirmed, per the company’s own age assurance documentation, is that the system ingests third-party data alongside platform-specific behavioral patterns.

There’s been wide speculation that sophisticated vocabulary or writing in full sentences causes the flag. The reasoning: the algorithm expects minors to write in a more abbreviated way, and adults who write carefully somehow look suspicious.

From what I’ve seen following this issue closely, that theory doesn’t hold up. Users who write formally are getting flagged and not flagged in roughly equal measure. Grammar doesn’t appear to be the deciding variable.

The more plausible read is that the system is doing population-level age estimation and has not been adequately trained to distinguish adult usage patterns from teen ones at the individual account level.

Character AI built this in response to high-profile lawsuits involving minors. The intent was right. The execution has been far messier than the company anticipated.

What Happens When You Get Flagged

The sequence is mostly consistent across affected users.

The seven-day lockout window

You’ll receive an in-app notice saying your access will be restricted within seven days. There may also be an email.

The message gives you the option to verify your age to prevent the lockout from taking effect.

What many users have run into is a timing gap: the warning arrives before the verification option is available on their specific account. They go to complete the process and find the feature isn’t there yet.

This has caused considerable panic among users who assumed they had missed a deadline.

Character AI has confirmed the rollout is staggered by account. If you don’t see the verification option immediately after the warning, it typically appears within a day or two.

The Persona verification steps

When the option becomes available, verification runs through Persona, a third-party company that provides identity checks for fintech platforms, banks, and technology services. Here’s the exact process:

1. Open the verification flow in the Character AI app or web interface
2. Persona prompts you to take a selfie in real time using your camera
3. Persona uses the selfie to estimate whether you’re over 18
4. If the selfie clears you, your account access is restored immediately
5. If the selfie fails to produce a clear result, Persona will request a government-issued ID
6. Character AI states it does not receive or store your ID. Persona sends only the pass or fail result back.
7. If you believe the system flagged you incorrectly, submit a support ticket to Character AI for manual review

The process is faster than it sounds. Most users who decide to verify clear it in under two minutes. The question isn’t whether it’s quick. The question is whether you want to do it at all.

What a flagged account notice looks like vs. a cleared one:

Flagged: You see a banner reading something like “Action required: your account will be restricted in 7 days. Verify your age to continue.” Features like message editing may already be disabled before you’ve taken any action.

Cleared: The banner disappears, and full feature access is restored. No ongoing monitoring or re-verification is required after a successful verification.

Does Good Grammar Really Trigger the Age Flag

Short answer: almost certainly not, and the persistence of this claim has muddied what’s a legitimate complaint about the system.

The claim spread quickly in user communities: writing in complete sentences or using a broad vocabulary triggers the algorithm, because it expects adult users to write in a more casual or abbreviated way and flags anyone who writes “too well” as potentially underage.

The problem is that this doesn’t match what’s happening. Adults who write formally are not getting flagged at a noticeably higher rate than those who write casually.

The system is generating false positives broadly, without a consistent grammatical pattern connecting them.

A more accurate framing: the algorithm is doing behavioral pattern matching and it’s calibrated poorly for individual accounts. It was trained on aggregate data.

When applied to any single user, it fails often enough to be a genuine problem for a large segment of the adult user base. That’s a model accuracy issue, not a vocabulary issue.

What the evidence shows:

Should You Hand Over a Selfie to Keep Your Account

This is the real decision point for most affected users, and it’s more personal than technical.

Persona is a legitimate company used by financial institutions and technology platforms worldwide. The data goes to them, not to Character AI.

The selfie you submit is processed by their age estimation system, and the ID upload only applies if the selfie is inconclusive. Per their stated policy, Character AI never sees any of it.

That said, submitting biometric data to any company is a decision you can’t walk back. Once a selfie is processed, it has been processed. Whether that matters depends on your personal threshold for privacy tradeoffs.

The more pointed question is whether Character AI is worth it at this point. The platform has been layering restrictions on its features for well over a year. It removed swipes and message playback. It capped conversation memory depth. It moved capabilities behind a charm economy that used to be free.

Now it’s asking users who have been on the platform since its early days to clear an identity hurdle to stay.

A lot of long-term adult users have looked at that list and decided the answer is no. That’s not an overreaction. It’s a reasonable response to a platform that has consistently traded user experience for compliance and cost-cutting.

Decision framework:

If you’re in the second or fourth column, the next section covers where to look.

Where Adults Are Going After Character AI

Most adults leaving Character AI aren’t stopping AI chat entirely. They’re moving to platforms that don’t require identity verification to access the features they want.

The two platforms worth trying first are Candy AI and Nectar AI. Both serve a similar user base to Character AI’s adult segment. Neither requires a biometric check to access adult content features. Both offer conversation capabilities that Character AI has increasingly restricted over the last year.

Candy AI is the better choice if you want to build characters from scratch with minimal friction on conversation topics.

The image generation features are worth noting if visual content was part of what you used Character AI for.

Nectar AI is the closer structural equivalent to Character AI’s companion model: longer memory, more personal interaction design, built around ongoing relationship-style interaction.

It won’t ask you to scan your face before letting you in.

For a broader breakdown of options across different use cases, the best Character AI alternatives with persona support covers more platforms in detail.

If the restrictions have been your main frustration, why Character AI’s filters are pushing users away is worth reading as context for what’s been building over the last year.