Kids safety on the internet is back on the agenda — with the usual caveat that nobody knows how it’s actually going to happen. A new wave of state-level bills is trying to put the onus on app stores rather than platforms like Meta — which could be more a recipe for a lobbying fight than real change. Congress is entertaining a new batch of proposals after the Kids Online Safety Act, a bill that tried to make social platforms safer for kids, failed to pass last year despite overwhelming support in the Senate. (The bill is poised to come back this year, but it’s unclear if it has a path forward without substantial changes.) A lot of the argument is based on who’s actually responsible for letting kids online, or liable when children have inappropriate interactions online. And the basic question of “how to do it” matters too. The Internet isn’t built to check ID’s. As the old joke goes, “On the Internet, nobody knows you’re a dog.” Any age-gating system runs quickly up against privacy concerns from both consumers and companies. This week, a new policy paper from a group close to the Trump administration proposed a high tech solution to the knotty problem of making the internet safe for kids without disrupting the adult experience. Taking advantage of cryptographic tools, blockchain technology, and artificial intelligence, companies could affirm someone’s age without storing additional data, policy experts with the Foundation for American Innovation write. The most effective place to do it, they say, is on the device itself. In their paper, FAI director of technology policy Luke Hogg and senior fellow Evan Swarztrauber say the way to make age verification policy impervious to legal challenges is to make it a requirement of operating systems rather than social platforms or apps. This way, age-verifying data, like an ID, can be stored or checked on a device and kept private. Operating systems can send off signals to apps on its platform, letting them know if a person is over or under 18. Doing so would make the requirement content neutral, they argue, meaning it has nothing to do with free speech. If it works, would there be a market for it? Maybe. There is already evidence that companies with websites or apps that children might visit are already starting to verify the age of their users and changing the experience for minors. Meta is rolling out age verification for Instagram in several different countries and just last week, Google said it was experimenting with using machine learning to verify a person’s age. App companies are also pushing for on-device age verification. In September, Pinterest CEO Bill Ready published an op-ed in Forbes advocating for an age-verification approach similar to the one FAI envisions. Meta CEO Mark Zuckerberg has also advocated for this approach. FAI’s policy ideas could also get a boost thanks to a close relationship with President Trump’s chief technology officer Michael Kratsios, who previously sat on the think tank’s board. Up to now, verifying kids' ages to provide them with more age appropriate experiences online has been a futile policy effort. The federal rule that is supposed to protect children online, the 1998 Children’s Online Privacy and Protection Act, requires affirmative consent from parents before collecting, using, or sharing data for kids under 13. But they only have to comply if they have actual knowledge of children using their platform, a huge loophole that has enabled sites to claim ignorance. Some 40 percent of 8-12 year olds use social media, according to a report from Surgeon General Vivek Murthy last year. FAI’s new paper suggests a way out of the policy boondoggle. It offers a variety of means for figuring out whether someone is over or under 18 years old. For example, advanced cryptographic tools could let users verify facts about themselves without giving up any actual data. Decentralized ledgers could verify the authenticity of credentials without recording them. AI could affirm someone’s age by scouring publicly available information. The approach might satisfy the adults grumbling about losing their ability to anonymously trawl the web. But device makers will likely push back. Apple already helped kill a provision in a Louisiana bill that would have saddled it with the responsibility of verifying kids’ ages. Whether or not policymakers take up the call, this movement to age-gate the internet is already happening. This year, YouTube will start experimenting with using machine learning to predict when a user is under 18 years old, and automatically restrict them from viewing content from sensitive categories like tobacco, alcohol, “dangerous activities” and weight loss. It will also turn on the highest privacy settings and stop collecting their location data. Others in the tech industry are also advocating for age verification. Ready’s September missive noted that parents already hand out their kid’s date of birth when they register a phone. He also highlighted that both Apple and Google, the leading smartphone companies, are working on developing digital state IDs. It seemed reasonable, he asserted, that a smartphone could share whether a person is a minor with the apps in its app store. Congress is already entertaining the idea. At a Senate Judiciary Committee hearing on online harms Wednesday, Sen. Ashley Moody (R-Fla.) asked witness John Pizurro, a police officer turned lobbyist who led an internet crimes division in New Jersey, what is the one policy Congress should focus on to protect kids online. “Device level age verification,” he said.
| 
