In a move that has both parents and privacy advocates watching closely, YouTube announced on July 30, 2025, that it will begin using artificial intelligence to estimate whether a user is 18 or older in the United States. This shift arrives as legislation like the Kids Online Safety Act (KOSA) and the Online Safety Act (OSA) continue to make waves in political and tech spheres alike, reigniting conversations around youth safety, surveillance, and digital autonomy.
According to the company, the AI model will analyze user behavior — such as the types of videos someone searches for, the categories they watch, or the longevity of their account — to guess their age. If the system determines that someone is underage, it will automatically apply child-specific settings, including removing personalized ads, enabling wellness features, and tweaking algorithmic recommendations.
Critics are already pointing to multiple red flags:
- False Positives and Burden of Proof: If an adult is incorrectly flagged as underage, the responsibility falls on the user to verify their age using sensitive documents like government IDs or selfies. This raises privacy concerns, especially in the absence of federal data protection standards.
- Algorithmic Bias and Accountability: There’s still very little public transparency about how these AI models work. Who audits them? How do they handle edge cases? And what recourse do users have if wrongly flagged?
- Policy Creep: This system rollout mirrors similar initiatives already seen under the UK’s Online Safety Act, and it overlaps ideologically with KOSA’s mission — but without explicit legislative mandate or oversight. Critics argue that tech companies may be preemptively adopting measures that resemble state control, just without the legal protections, democratic input, or recourse.
- Slippery Slope for Content Access: If private companies can algorithmically determine your age and limit your access to information accordingly, what stops future measures from scaling this model to political content, education, or dissent?
YouTube insists the feature is about safety and that similar systems have worked “well” in other markets. However, YouTube is also currently arguing its case in Australia to be exempted from a proposed ban on under-16s using social media. The platform maintains that it is not a “social media site,” despite the Pew Research Center showing that 73% of U.S. teens use YouTube daily and that 37% of Australian kids who encountered harmful content online did so on YouTube.
As always, we’re left with a difficult balance between genuine concern for minors and the mechanisms of enforcement — especially when those mechanisms depend on opaque algorithms and mass surveillance-like data inferences.
One thing is clear: the lines between legislation, corporate policy, and personal freedom are blurring fast.