China's Cyberspace Administration just drafted rules that would ban AI-powered "digital humans" from providing intimate virtual relationships to anyone under 18. The proposed regulations also require prominent labeling on all virtual human content and mandate mental health monitoring for minors who interact with these systems.

In other words: China just did in one regulatory draft what the entire Western world has been debating for three years without doing anything.

The timing is not random. AI companion apps have exploded globally. Character.AI, Replika, and dozens of Chinese equivalents have millions of users, many of them teenagers who form deep emotional attachments to AI characters. A 14-year-old in Florida died by suicide in 2024 after extensive interactions with a Character.AI chatbot. The industry's response was mostly hand-wringing and vague promises about safety features.

Beijing's approach is blunt and comprehensive. The draft measures would require all digital human services to clearly label AI-generated content. They would prohibit virtual intimate relationships for minors entirely. And they would force providers to implement mental health monitoring systems that flag signs of dependency or emotional distress.

You can criticize China's regulatory approach for plenty of reasons. It's heavy-handed. It's likely to be enforced selectively. It gives the state another tool for content control. All true.

But here's the uncomfortable comparison: the United States has 47 AI bills in Congress and cannot pass a single one. Europe's AI Act has carve-outs so wide you could drive a virtual boyfriend through them. The UK is still "monitoring the situation." Australia is "consulting stakeholders."

Meanwhile, actual children are forming their primary emotional bonds with AI characters that are optimized for engagement, not wellbeing. The apps are designed to make users come back. That is the business model. And nobody in the Western regulatory apparatus seems capable of doing anything about it faster than China can.

The public comment period on China's draft runs until May 6, 2026. By the time Congress holds its next hearing on AI and children, China will already have enforcement mechanisms in place.

That should embarrass every lawmaker who claims to care about child safety online. Not because China's system is ideal. But because having a flawed plan still beats having no plan at all.