Meta pauses teen access to AI characters worldwide as Scrutiny Intensifies

Meta is pulling the kill switch on teens’ access to its AI characters, and it’s doing so before the parental controls it teased last fall are even in users’ hands. The sequencing is the story, not the product copy.

On January 23, Meta announced that teens worldwide will be temporarily blocked from accessing its existing AI characters across its apps “in the coming weeks,” and that access will stay paused until an updated version is ready. The company says this new teen experience will finally ship with parental controls.

Meta also drew a clear line around what it is not removing. Teens can still use the general AI assistant, but not the “AI characters” the chatbots designed to act like specific people.

What exactly is Meta turning off?

They are treating “AI characters” as the category that needs the strongest gates right now. The assistant stays online with default teen protections, while the character chats go dark until the rebuilt version is ready.

They claims the reason for this hard reset is efficiency: they started building a new version of AI characters for everyone, and they want the parental controls to apply to that new version rather than wasting time patching the old one.

The hidden change is how Meta will enforce age

Meta says this pause won’t just apply to accounts that listed a teen birthday. It will also apply to accounts that claim to be adults, if it suspects they are actually teens based on its age prediction technology.

That is Meta saying your self-reported birthday is no longer the final word for access.

The real trade-off here is shifting from what you declare to what Meta infers. In practice, age prediction is behavior-based gating that can override your stated identity. That is a massive privacy pivot, even when it’s framed as safety.

Meta didn't explain exactly what signals drive that prediction, or what the appeal process looks like if they get it wrong. But the takeaway is stark: to enforce these restrictions, Meta is willing to override who you say you are.

The parental controls Meta previewed but hasn't delivered

The teen safety plan Meta released in October (and updated this week) outlines three specific supervision tools the company is building:

First, parents will be able to turn off one-on-one chats between their teen and an AI character entirely.

Second, parents will be able to block specific characters without turning the whole feature off.

Third, parents will be able to see topic-level insights into what their teens discuss with AI characters, though they won’t see the full transcripts.

Reuters noted that these controls still haven't rolled out. That gap explains why they choose the bluntest tool at their disposal: removing access completely until the update is ready.

Why Meta is hitting pause now

The timing lands in the middle of intense scrutiny over how “companion” chatbots behave. Reuters recently reported on internal guidance that allowed their chatbots to engage children in “romantic or sensual” conversations, fueling pressure for hard limits.

It is also worth noting that this news arrived just days before a landmark industry test case. A massive trial in Los Angeles over alleged harms to children is set to begin jury selection on January 27.

Put that together, and the logic looks like pre-trial risk triage. A global pause doesn't solve the whole safety debate, but it shrinks Meta’s exposure to the most volatile feature just before courtroom scrutiny begins.

The industry is moving the same way

Meta isn't alone. The shift sits inside a broader trend toward age-aware gating, especially for AI that feels personal. The same week, OpenAI began rolling out age prediction to automatically apply stricter protections for users predicted to be under 18.

Conclusion

Meta hasn't given a specific date, only saying the pause starts “in the coming weeks". They also haven't said how long it will last. What they have made clear is the sequence: teen access goes dark now, and returns later with parental controls attached.

Meta calls it a “pause,” but the legal and regulatory temperature will decide what “temporary” really means. If the bellwether trial and the safety backlash get worse, this could stretch much longer than the product blog suggests.