Meta Deploys AI Body Scans to Detect Underage Users

Meta is rolling out a new AI-powered visual analysis system that scans users' physical characteristics — including height and bone structure — to determine whether they are underage. The system is already operational in select countries, with the company confirming plans for a broader global rollout in the coming months.

The move represents one of the most aggressive and technically ambitious approaches any major social media platform has taken to enforce age restrictions, raising both praise from child safety advocates and sharp questions from privacy experts.

Key Takeaways at a Glance

• Meta's new system uses computer vision AI to analyze height, bone structure, and other physical markers to estimate a user's age
• The technology is currently live in select countries, with a wider rollout planned
• Unlike traditional age verification methods like ID uploads or self-reported birthdates, this approach requires no documents
• The system adds to Meta's growing suite of youth protection tools, including Teen Accounts on Instagram launched in late 2024
• Privacy advocates have raised concerns about biometric data collection and potential biases in the AI model
• Meta says the analysis happens in real time and that raw visual data is not stored long-term

How Meta's Age-Detection AI Actually Works

Meta's new system relies on computer vision models trained to estimate biological age from visual cues. When a user's age is in question — for instance, if their behavior patterns or account metadata suggest they may be younger than claimed — the system can trigger a visual analysis.

The AI examines physical characteristics such as facial bone structure, body proportions, and estimated height to generate an age estimate. This is fundamentally different from simple facial recognition. Rather than identifying who someone is, the system attempts to determine how old they are based on developmental markers.

According to Meta, the technology draws on research in forensic anthropology and pediatric growth modeling. Human bone structure, particularly in the face and jaw, undergoes predictable changes during adolescence. By training deep learning models on large datasets of age-verified images, Meta says it can estimate whether a user falls below the platform's minimum age threshold of 13.

The system reportedly works alongside other signals. Meta has previously described using behavioral analysis — such as the types of accounts a user follows, messaging patterns, and content engagement — as supplementary age indicators. The visual analysis adds a biometric layer to this multi-signal approach.

Why Meta Is Going Beyond Traditional Age Verification

Traditional age gates on social media have been widely criticized as ineffective. Self-reported birthdates are trivially easy to fake — a child simply enters a false year of birth and gains unrestricted access. Even government ID-based verification, which platforms like Pornhub and some state-mandated systems use, has significant drawbacks.

• ID verification excludes users without government-issued documents
• Self-reported ages are easily falsified by minors
• Parental consent flows are often bypassed by tech-savvy children
• Credit card verification creates friction and raises its own privacy concerns

Meta's biometric approach attempts to sidestep these issues entirely. Users don't need to upload documents or take any explicit verification step. The AI works passively, analyzing available visual data — such as profile photos or video content — to flag accounts that may belong to underage individuals.

This passive approach is key to Meta's strategy. The company has faced mounting regulatory pressure worldwide, from the EU's Digital Services Act to proposed legislation in the United States like the Kids Online Safety Act (KOSA). Australia has gone even further, passing a law in late 2024 that bans children under 16 from social media entirely. Meta needs scalable solutions, and AI-driven age estimation offers exactly that.

Privacy Concerns and Ethical Questions Mount

Not everyone is applauding the move. Digital rights organizations have raised serious concerns about the implications of AI systems that analyze users' bodies, even for ostensibly protective purposes.

The core worry is about biometric data collection. Height estimation and bone structure analysis are, by definition, biometric processes. In jurisdictions like the European Union, biometric data receives the highest level of legal protection under the General Data Protection Regulation (GDPR). Illinois' Biometric Information Privacy Act (BIPA) has already been the basis for billions of dollars in settlements against tech companies, including a $1.4 billion settlement Meta paid in 2024.

Critics also point to potential algorithmic bias. Age estimation models trained primarily on certain demographic groups may perform poorly on others. Children from different ethnic backgrounds, or those with medical conditions affecting growth, could be misclassified. A false positive — incorrectly flagging an adult as underage — could lock legitimate users out of platform features. A false negative could leave vulnerable children exposed.

• Data retention: How long does Meta store the visual data used for analysis?
• Consent: Are users informed before the AI scans their physical characteristics?
• Accuracy: What is the system's error rate across different demographics?
• Appeals: Can users challenge an incorrect age determination?
• Scope creep: Could this technology be repurposed for other forms of surveillance?

Meta has stated that raw visual data is not retained after analysis and that the system produces only an age estimate, not a biometric template. However, independent audits of these claims have not yet been published.

How This Compares to Industry Rivals' Approaches

Meta is not the only company investing in AI-powered age estimation, but its approach is arguably the most invasive. Yoti, a UK-based digital identity company, offers facial age estimation technology that several platforms already use. Yoti's system analyzes facial features from a single selfie and claims accuracy within 1.5 years for users aged 13 to 17.

Google requires age verification for certain YouTube content but primarily relies on ID uploads or credit card checks. TikTok, which has faced intense scrutiny over youth safety, uses a combination of self-reported ages and AI-driven behavioral analysis but has not publicly deployed body-scanning technology.

Apple takes a different approach entirely, embedding parental controls at the operating system level through Screen Time and Family Sharing, rather than relying on platform-level age detection.

Meta's decision to incorporate bone structure and height analysis goes significantly beyond facial age estimation. It suggests the company is building a more comprehensive biometric profile, even if temporarily, to make its age determinations. This positions Meta at the leading edge of the technology — but also at the frontier of the privacy debate.

Regulatory Landscape Pushes Tech Giants to Act

The timing of Meta's deployment is no coincidence. Governments around the world are tightening requirements for platforms to protect minors, and the penalties for non-compliance are escalating.

In the United States, the Federal Trade Commission (FTC) has proposed sweeping updates to the Children's Online Privacy Protection Act (COPPA), which would impose stricter requirements on how platforms handle data from users under 13. Several US states, including Utah, Texas, and California, have passed or proposed their own youth safety laws.

The European Union's Digital Services Act requires platforms to assess and mitigate risks to minors, with potential fines of up to 6% of global annual revenue for violations. For Meta, which reported approximately $135 billion in revenue in 2024, that could mean penalties exceeding $8 billion.

Australia's outright ban on social media for children under 16, signed into law in November 2024, puts additional pressure on platforms to develop reliable age verification. Meta's AI-based approach could serve as a technical solution to comply with such mandates at scale.

What This Means for Users, Parents, and Developers

For everyday users, the most immediate impact may be invisible. Meta's system appears designed to operate in the background, analyzing visual data without requiring active participation. Adults who use the platform normally should notice no difference.

For parents, the technology represents a potentially meaningful layer of protection. If the AI can reliably identify and restrict underage accounts, it could reduce children's exposure to harmful content, predatory contacts, and addictive platform features. However, parents should remain aware that no automated system is perfect.

For developers and the broader tech industry, Meta's deployment sets a precedent. If this approach proves effective and legally defensible, other platforms will face pressure to adopt similar technologies. AI age estimation could become a standard component of platform safety infrastructure within the next 2 to 3 years.

Looking Ahead: A New Standard or a Step Too Far?

Meta's AI body-scanning system sits at the intersection of child safety and digital privacy — 2 values that increasingly collide in the modern internet. The company is betting that the public's appetite for protecting children will outweigh concerns about biometric surveillance.

The broader rollout, expected over the course of 2025, will be a critical test. Regulators, advocacy groups, and independent researchers will scrutinize the system's accuracy, fairness, and data practices. If Meta can demonstrate high accuracy across demographics and transparent data handling, the technology could become an industry benchmark.

But if the system proves biased, error-prone, or opaque, it could fuel the growing backlash against Big Tech's expanding surveillance capabilities. The stakes are high — not just for Meta, but for the future of how platforms verify identity in an increasingly regulated digital world.

One thing is certain: the era of relying on a simple birthdate dropdown menu for age verification is coming to an end. What replaces it will define the next chapter of online safety — and online privacy — for billions of users worldwide.