Oct 17 (Reuters) - Meta said on Friday it will

let parents disable their teens' private chats with AI

characters, adding another measure to make its social media

platforms safe for minors after fierce criticism over the

behavior of its flirty chatbots.

Earlier this week, the company said its AI experiences for

teens will be guided by the PG-13 movie rating system, as it

looks to prevent minors from accessing inappropriate content.

U.S. regulators have stepped up scrutiny of AI companies

over the potential negative impacts of chatbots. In August,

Reuters reported how Meta's AI rules allowed provocative

conversations with minors.

The new tools, detailed by Instagram head Adam Mosseri and

Chief AI Officer Alexandr Wang, will debut on Instagram early

next year, in the U.S., United Kingdom, Canada and Australia,

according to a blog post.

Meta said parents will also be able to block specific AI

characters and see broad topics their teens discuss with

chatbots and Meta's AI assistant, without turning off AI access

entirely.

Its AI assistant will remain available with age-appropriate

defaults even if parents disable teens' one-on-one chats with AI

characters, Meta said.

The supervision features are built on protections already

applied to teen accounts, the company said, adding that it uses

AI signals to place suspected teens into protection even if they

say they are adults.

A report in September showed that many safety features Meta

has implemented on Instagram over the years do not work well or

exist.

Meta said its AI characters are designed not to engage in

age-inappropriate discussions about self-harm, suicide or

disordered eating with teens.

Last month, OpenAI rolled out parental controls for ChatGPT

on the web and mobile, following a lawsuit by the parents of a

teen who died by suicide after the startup's chatbot allegedly

coached him on methods of self-harm.