Part of the challenge lies in the inherent difficulty of teaching algorithms to make nuanced decisions about content
Thomas Trutschel/Photothek via Getty Images
A version of this article originally appeared in Quartz’s members-only Weekend Brief newsletter. Quartz members get access to exclusive newsletters and more. Sign up here.
When British lawmakers promised that age-verification laws would create a “safer internet” by keeping children away from harmful or adult content, they likely didn't envision adults needing government ID to view parliamentary debates about sexual abuse or footage of destroyed drones in Ukraine.
Join 500,000+ readers who start their day with Quartz.
By subscribing, you agree to our Terms of Service and Privacy Policy.
But that's exactly what's happening as countries across the world rush to implement age checks across the web. The gap between policy intentions and messy reality reveals the fundamental challenge of creating technology sophisticated enough to distinguish between content that's genuinely harmful to children and legitimate information that adults have every right to access.
The consequences are already rippling across the internet. In the U.K., where the Online Safety Act took effect in late July, social media platforms are indiscriminately blocking content in their scramble to comply with new regulations. A video of a man in Gaza searching for his family's bodies among rubble was restricted on X $TWTR, despite not showing any graphic imagery, until the BBC inquired about it.
A parliamentary speech about a sexual assault available without restriction on the government's official streaming site is blocked on social media. The algorithms have even deemed a 200-year-old masterpiece depicting Greek mythology (Goya's "Saturn Devouring His Son") as too spicy for the internet. Meanwhile, downloads of VPN apps have surged as users seek workarounds, a clear signal that the public isn't embracing this new reality.
Part of the problem lies in the inherent difficulty of teaching algorithms to make nuanced decisions about content. The U.K.'s Online Safety Act requires platforms to assess whether content presents "a material risk of significant harm to an appreciable number of children," a subjective determination that even humans struggle with consistently.
The challenge compounds when platforms try to determine who should be blocked from seeing that content in the first place. Companies are outsourcing these decisions to third-party age verification providers such as Yoti, Incode, and Persona, but the technology remains imperfect. Facial age estimation systems typically offer accuracy within roughly 18 months, and can be fooled by everything from poor lighting to, in one case, a dog held in front of someone's face.
The verification methods themselves create new problems. Requiring government IDs raises privacy concerns and excludes people without documentation. Credit card checks assume financial access. Even seemingly privacy-friendly approaches, like AI age estimation based on browsing patterns, struggle to account for shared devices or varied online behaviors.
But perhaps most concerning is what happens when users simply opt out. Data suggests the laws may be pushing users toward less regulated sites. When age verification requirements take effect, researchers have found that traffic often shifts to overseas, unregulated sites that don't follow any content policies.
Despite these challenges, the age verification experiment is expanding rapidly. The U.S. Supreme Court recently paved the way for states to require porn websites to check visitors' ages, with at least 24 states having enacted some form of verification law. Australia plans to ban social media for children under 16 starting in December. France's courts have ruled that porn sites can check users' ages, while Ireland implemented age-checking laws for video websites earlier this year.
Each jurisdiction is taking a slightly different approach, creating inconsistent regulatory frameworks of compliance requirements that platforms must navigate. Texas fines non-compliant sites up to $10,000 per day, while the U.K. can fine platforms up to 10% of global revenue or block services entirely.
Tech trade associations are pushing back in court, with NetChoice — whose members include Meta $META, Google $GOOGL, and Reddit $RDDT — arguing that the laws are counterproductive since determined young users will simply find ways around the restrictions.
Researchers worry about the broader implications for information access. For adults, age verification requirements eliminate anonymous browsing. This concern extends beyond pornography to include politically sensitive topics, mental health resources, or any content someone might legitimately want to access privately.
For young people, the laws may inadvertently restrict access to educational materials they need most, including information about sexual health, reproductive rights, and LGBTQ issues that could be vital in conservative jurisdictions where such topics face political opposition.
As more countries implement these laws, the internet is becoming a more fragmented, surveilled space. The fundamental tension may prove irreconcilable: Lawmakers have created a system that blocks parliamentary debates while failing to keep determined teenagers away from the content it was designed to restrict.
Whether this represents progress in child safety or an expensive lesson in the limits of technological solutions remains to be seen.
