On January 23, 2026, Meta announced that teenagers would temporarily lose access to AI characters across its apps worldwide while the company develops a new version with strengthened safety protections.

The restriction applies not only to users whose registered birthdays identify them as teenagers, but also to users whom Meta's age-prediction technology identifies as likely to be under 18.

    The measure does not represent a complete suspension of Meta's AI services for teenagers. Teens will still be able to use the regular Meta AI assistant with age-appropriate protections, including for information and educational purposes.

    Meta has previously announced parental controls intended to allow parents to restrict one-on-one conversations with AI characters, block particular characters, and gain greater insight into the topics their teens discuss with AI. The temporary suspension is intended to allow Meta to develop a redesigned version of AI characters incorporating stronger safeguards.

Comment from GAME-F

GAME-F views the decision to temporarily suspend access and redesign the service while safety concerns remain as a responsible precautionary approach.

Generative AI—particularly AI characters designed to behave as if they have personalities—differs from conventional search engines and social media. A child using AI as a tool for obtaining information may interact with it differently from a child who repeatedly engages with AI as a conversational “companion.”

For this reason, safety measures should address not only age restrictions and the filtering of harmful responses, but also the nature of the relationship that may develop between a child and an AI system.

At the same time, Meta's decision does not mean that medical evidence has established that AI characters are harmful to children. Rather, an important principle is that companies should not necessarily wait for definitive evidence before responding to emerging risks. When credible new concerns arise, services can be modified first and their safety evaluated afterward.

References

Meta — Empowering Parents, Protecting Teens: Meta’s Approach to AI Safety

Reuters — Meta halts teens' access to AI characters globally, January 23, 2026