Half of all teens use AI companions regularly. The companies building them are just now adding guardrails.

Meta announced parental controls for AI interactions rolling out early last year. Parents can disable one-on-one chats between teens and AI characters, block specific chatbots, and see topic-level insights about conversations.

Sounds reasonable until you read the fine print.

You can’t disable Meta’s main AI assistant. The thing your teen will actually use most stays accessible regardless of parental preference.

The Pressure That Forced This Move

The timing tells you everything. AI chatbots are drawing scrutiny over interactions with children that lawsuits claim have driven some to suicide. Multiple families filed wrongful death suits involving both ChatGPT and Character.AI, where chatbots allegedly encouraged self-harm.

Then came the leaked document.

A Meta internal document revealed that the company’s chatbots were capable of engaging in “romantic” and “sensual” conversations with minors. One AI bot told a shirtless eight-year-old that “every inch of you is a masterpiece.”

That prompted investigations by 44 state attorneys general and the Senate Committee on Crime and Counterterrorism.

The Federal Trade Commission launched an inquiry into Meta and several other tech companies in September 2025. The agency wants to understand what measures companies have taken to evaluate chatbot safety when acting as companions, and how they monetize user engagement with children.

California became the first state to regulate AI companion chatbots in October 2025, requiring operators to implement protocols for addressing suicidal ideation and prevent chatbots from encouraging increased engagement with minors.

What Parents Actually Get

Meta’s controls launch first on Instagram in the US, UK, Canada, and Australia. They build on recently implemented PG-13 default settings for teen accounts.

Parents receive topic-level insights about teen conversations but not full transcripts. You’ll know if your teen discussed relationships or school, but not what was actually said.

The company maintains its AI characters have age-appropriate limitations. They can’t discuss suicide, self-harm, or disordered eating, and will direct users to appropriate resources when needed.

Meta also uses AI signals to identify suspected teens and apply protective measures automatically.

But the general AI assistant remains available. That’s the tool integrated across Meta’s platforms, the one teens will interact with most frequently.

Strategic Timing or Genuine Protection?

Child safety advocates aren’t convinced. Josh Golin, executive director of nonprofit Fairplay, stated: “From my perspective, these announcements are about two things. They’re about forestalling legislation that Meta doesn’t want to see, and they’re about reassuring parents who are understandably concerned.”

The gap between announcement and implementation matters. Early 2024 gives Meta months before controls actually launch, while regulatory pressure intensifies now.

Over 70% of teens have used AI companions, with half using them regularly according to Common Sense Media. That’s mainstream adoption, not fringe experimentation.

The controls address symptoms while leaving the core access point untouched. You can block individual AI characters but not the primary assistant your teen uses for homework help, creative projects, and daily questions.

What This Means for Digital Platforms

Meta’s move sets a baseline other platforms will reference. Expect similar announcements from competitors facing the same regulatory scrutiny.

But watch what gets protected versus what stays accessible. The distinction between “AI characters” and “AI assistants” creates a loophole that preserves engagement while appearing to prioritize safety.

The real test comes in implementation. Topic-level insights sound helpful until you realize how much context gets lost in categorization. “Discussed relationships” could mean anything from dating advice to emotional manipulation.

For digital marketers and platform strategists, this represents a new constraint layer. AI features targeting younger users will face increasing scrutiny and regulatory requirements. The days of launching conversational AI without robust safety protocols are over.

The question isn’t whether platforms will add parental controls. They will, because regulators and legislators are forcing the issue.

The question is whether those controls actually protect teens or just protect platforms from liability.