Meta will use AI to analyze height and bone structure to identify if users are underage

Meta's decision to deploy AI systems capable of analyzing physical attributes like height and bone structure to determine user age represents a significant escalation in the technology industry's approach to underage verification. While the company frames this as a child safety measure, the implications extend far beyond protecting minors online. This approach raises fundamental questions about privacy, consent, and the appropriateness of using biometric analysis at scale.

The technical foundation underlying this system warrants scrutiny. Physical characteristics like height and bone structure vary dramatically across populations, and relying on visual analysis introduces substantial error rates that could disproportionately affect certain user groups. Moreover, the notion that AI can accurately determine age from physical appearance conflates correlation with causation and overlooks the complex reality of human development. A system's ability to make such determinations with sufficient accuracy for mass deployment remains scientifically questionable, yet Meta is moving forward regardless. For those tracking the evolution of AI systems in production environments, this raises familiar concerns about evaluation rigor and the risks of deploying technology before it reaches maturity. The conversation around building reliable AI evaluation frameworks, explored in discussions like Building an Evaluation Harness for Production AI Agents: A 12-Metric Framework From 100+ Deployments, remains critically relevant as companies push AI into increasingly sensitive applications.

Beyond technical accuracy, this move signals a troubling normalization of biometric surveillance as a default solution to platform safety challenges. Rather than exploring less invasive verification methods or addressing the underlying design choices that attract underage users to certain platforms, Meta has opted for an approach that treats all users as potential subjects of physical analysis. The cumulative effect of such systems is an internet where physical bodies become subject to algorithmic scrutiny, often without meaningful consent or transparency about how such data is used and stored. This represents a broader shift in the relationship between platforms and users, one where the burden of proof increasingly shifts to individuals to prove they deserve privacy rather than platforms demonstrating why they need to collect intimate data.

The regulatory landscape remains unprepared for this reality. Existing frameworks around the world were designed with traditional data collection in mind, not AI-driven physical analysis systems. The European Union's upcoming enforcement of the Digital Services Act may provide some constraints, but global platforms operating across jurisdictions with varying standards create a fragmented compliance environment that often defaults to the lowest common denominator. Users in countries with weaker privacy protections will find themselves subject to biometric analysis that users in more protective jurisdictions might successfully resist.

What should concern observers most is the trajectory this establishes. If physical analysis becomes accepted for age verification, what prevents similar technology from being deployed for content moderation, targeted advertising, or other commercial purposes? The normalization of biometric analysis happens incrementally, and each deployment makes the next seem more reasonable. The question facing users, regulators, and the broader technology ecosystem is whether we want a future where physical appearance becomes just another data point in the vast surveillance apparatus of platforms, or whether meaningful boundaries exist around what companies can analyze without explicit, informed consent.