Kenya weighs child social media curbs as Meta age checks go global

A Kenyan teenager seeking to access Instagram may enter an adult birth date during registration to avoid restrictions designed for younger users.

But now, that trick is becoming harder to pull off as social media companies turn to artificial intelligence (AI) to identify underage users.

Meta, the owner of Facebook, Instagram, Messenger and WhatsApp, is expanding AI-powered age-estimation technology globally to identify teenagers who may have registered on its platforms using false dates of birth.

The technology is designed to detect users under 17 even when their profiles indicate they are adults, automatically moving them into accounts with additional safety restrictions.

It marks a shift in how tech companies enforce age limits, moving beyond relying on information supplied by users to analysing their online activity, profile content and other signals that may indicate their age.

For Kenya, the expansion comes as Parliament considers legislation that would require social media companies to obtain parental consent before allowing children to access their platforms, and deploy age-verification systems.

Meta says its technology uses AI to examine contextual clues in accounts, including references to school grades and birthday celebrations, as well as visual cues in photographs and videos.

The company says the visual analysis does not involve facial recognition.

‘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 company said on Monday. Teen Accounts apply stricter privacy, messaging and content controls.

‘Since first introducing this technology, we’ve been placing more teens in age-appropriate protections and are encouraged by the results. Based on that success, today we’re expanding this technology globally.’

Meta is also extending the system to Facebook and plans to introduce similar age-detection technology on Threads before the end of the year.

Teen Accounts are designed for users aged 13 to 17 across Instagram, Facebook and Messenger, three of the most-used social media platforms in Kenya.

These accounts automatically default to private settings, restricting who can follow, message, tag or mention teenagers.

They also limit exposure to content related to self-harm and graphic subjects, and introduce tools to manage time spent on the platforms.

Teenagers receive reminders after 60 minutes of daily use, for instance, and have a ‘sleep mode’ which silences notifications between 10pm and 7am.

Users under 16 are also barred from live-streaming without parental permission.

Through the parental supervision tool Family Centre, parents and guardians can monitor aspects of their children’s online activity, set time limits and scheduled breaks, see the accounts their teenagers follow or interact with and receive certain safety alerts.

They cannot, however, read the contents of private messages.

The global expansion reflects growing pressure on social media companies to demonstrate that they can identify minors and protect them from harmful online experiences.

Meta has faced legal and regulatory scrutiny in the United States over allegations that its platforms contribute to addictive behaviour among children, expose them to harmful content and collect their personal information without adequate safeguards.

Governments are also increasingly considering age restrictions, parental consent requirements and stronger obligations for platforms to protect minors.

In Kenya, the proposed Online Safety Bill 2026 seeks to make parental consent a condition for children accessing regulated online services, including Facebook, Instagram, WhatsApp, X and TikTok.

The proposal requires platforms to establish whether a user is a child through age-verification or age-estimation technologies and obtain consent from a parent or another person exercising parental responsibility before granting access.

It also sets out ways platforms could verify parental responsibility, including credit card checks, cross-referencing a parent’s identity and decentralised digital identity verification, alongside other methods approved by the Communications Authority of Kenya.

Parents would be able to withdraw consent, requiring platforms to restrict the child’s access within a period prescribed under the law.

Categories of content considered harmful to children include pornography, material encouraging suicide or deliberate self-harm, content promoting eating disorders and dangerous online challenges.

Beyond the tech companies, the proposal assigns parents and guardians responsibility for supervising children’s online activities according to their age and maturity, and reporting cases of abuse or exploitation.

If enacted, companies that fail to meet the proposed requirements will face fines of up to Sh500 million or 10 percent of their annual global turnover, whichever is higher.

Still, using AI to estimate someone’s age has also raised questions about privacy, accuracy and the potential exclusion of legitimate adult users.

Meta’s approach involves analysing information and activity on accounts, meaning age estimation could draw on a range of personal data beyond the birth date supplied during registration.

Privacy advocates have raised concerns about how such information is processed, retained and independently audited.

There is also the risk of false positives, where adults are incorrectly identified as teenagers and placed under restrictions, or false negatives, where underage users evade detection.

For parents, the changes could provide another layer of protection and greater visibility into how children use social media.

For businesses, the requirements could affect digital advertising and marketing businesses that rely on reaching younger audiences.

Stronger age controls may limit the data available for profiling minors and restrict how platforms target them with advertising and other commercial content.

Leave a Reply

Your email address will not be published. Required fields are marked *