Meta rolls out AI age checks for teens, starting with Instagram worldwide

Meta is rolling out automated age-estimation technology worldwide to better protect teenagers. Artificial intelligence will scan Instagram accounts to find users suspected of being teens, even if they signed up using an adult birthday.

In an announcement on Monday, the company said the AI-powered checks will begin on Instagram this month and will later be extended to Facebook and Threads.

Meta said: "We use AI technology to find people we suspect to be teens, even if they use an adult birthday, and place them in Teen Account protections."

The technology analyses a user's behaviour and posts to estimate their age. It had been tested in several countries, including the United States, the United Kingdom, Canada and Australia; based on that success, it is now being expanded globally.

Teen accounts carry more restrictions and protections. Children under 13 are banned from using the apps. Teenagers are reminded to leave Instagram after 60 minutes of activity, and notifications are muted at night, when they would usually be sleeping.

Teen accounts also face a two-hour-a-day usage limit and are blocked from seeing "social comparison" features such as likes and reactions to other teen accounts.

Parents in the United States who have set up Instagram for teenagers will receive notifications this month about the new measures, Meta said.

The social media giant has faced growing legal action worldwide over how its platforms are used by teenagers. Child development experts have said platforms like Instagram and Facebook are highly addictive for young people, can reinforce social pressures and expose them to other harmful content.

Australia passed a ban last year on anyone below 16 years of age from using social media platforms. But this month, several groups, including the advocacy group Digital Freedom Project and the social media website Reddit, filed a lawsuit challenging the constitutionality of that law, as well as the scientific rationale behind trying to keep children entirely off of platforms like YouTube, Facebook and TikTok.

In the US, Meta faced lawsuits from attorneys general across the country over the alleged harmful impact of its platforms on children. The states alleged that Meta's platforms "encourage addictive behaviour, fail to verify users' ages, encourage adolescents to bypass parental controls, and inadequately safeguard against harmful content."

In August, the company agreed to pay $18bn to the states and institute a slew of changes to how it allows teenagers to interact with its platforms. The changes include restricting under-18 users to two hours a day on the platforms and enforcing a nighttime curfew from midnight to 6am. Default settings will restrict "social comparison" features such as seeing likes and reactions on posts, and "cosmetic procedure filters" that alter a person's appearance. Push notifications will be blocked for teen accounts between 8am and 3pm, when they are likely to be in school.

Teen users will be encouraged to take breaks, with prompts appearing after every 15 minutes of continuous usage, and when daily usage hits 60 minutes and 90 minutes. Parents and guardians will receive updates on teen account activities and be notified whenever they message an adult for the first time. Searches for topics like suicide, self-harm or eating disorders would also trigger a notification to parents.

In announcing the new automated age estimation on Monday, Meta also repeated calls for other industry leaders to implement similar changes. "As Instagram is just one of the many apps teens use, we're also calling on our industry peers, TikTok and YouTube, to adopt these same practices to keep teens safe," Meta said in its statement.