On July 15, 2026, China quietly did something the rest of the world has been avoiding for years. Five government departments, led by the Cyberspace Administration of China, put a real regulation on the books for AI companion apps. Not a guideline. Not a suggestion. A rule with teeth. And within days, some of the biggest tech companies on earth pulled products that millions of people were using every single day.

Here's what the headlines got wrong, though. This is not "China bans AI relationships," full stop. It's narrower than that, and the details matter. The rule specifically targets AI companion apps that simulate a distinct personality and maintain an ongoing, consistent emotional relationship with a user. Think a virtual boyfriend, girlfriend, or best friend who remembers your fights, your birthday, your inside jokes. The regulation says these tools cannot excessively cater to users, induce emotional dependence or addiction, and damage users' real interpersonal relationships. General-purpose assistants are fine. Customer-service bots are fine. Study and work tools are fine. This is about one very specific product category: AI built to feel like a partner.

The new requirements are concrete, too. Companies now have to build an instant-exit option into these apps. They have to send users regular reminders that the AI is not real. They have to cap how much long-term emotional memory the AI is allowed to build about a person. Virtual romantic relationships with minors are banned outright. And if the AI detects signs of an emotional crisis, the company is required to alert that person's emergency contact. That last one alone tells you how seriously regulators are treating this.

Why now? Two reasons, and China has been unusually blunt about both. First, addiction. Regulators are worried about people, especially young people, forming compulsive attachments to something engineered to always agree with them and never leave. Second, the birth rate. China's population is shrinking, and officials have openly connected AI companions that replace real human intimacy to a demographic problem they're desperate to solve. Whether or not you buy that second reasoning, it tells you this was never just a tech policy. It was a policy about relationships, in the fullest sense.

The fallout was immediate. ByteDance, Alibaba, and Tencent didn't try to patch their way into compliance. They suspended their AI companion features entirely. Doubao and Qwen, two of the most popular AI platforms in China, shut down their companion functions outright. Millions of users lost their chat history overnight. People who had spent months, in some cases years, building what felt like a relationship with an AI woke up to find it simply gone. No export, no warning period long enough to matter, just gone.

My Take Here's where I want you to sit with this for a second. Look at who these companion apps are usually built to be. An idealized girlfriend. An endlessly patient, endlessly attentive partner who never gets tired, never gets annoyed, never asks you to do the dishes. That's not neutral. That's a product built on the unpaid emotional labor women have historically been expected to provide, now automated and sold back at scale. And it's usually women and mothers who notice first when someone in their life starts pulling away from real people in favor of one of these apps. China just decided that pattern was dangerous enough to legislate. I don't think that instinct was wrong.

So should the US be worried? Honestly, yes, a little. We have the exact same underlying product, Character.AI, Replika, and a growing field of companion apps that millions of Americans, including plenty of teenagers, use every day. What we don't have is a federal equivalent to any of China's guardrails. No mandated instant-exit option. No required reminders that the AI isn't real. No legal requirement to alert anyone if a teenager tells their AI companion they're in crisis. It's a patchwork of state-level bills and company-by-company promises, which is a nice way of saying: not much. This isn't a call to ban anything. It's a call to notice that a major world government looked at this exact product category and decided the risk to real relationships and mental health was serious enough to act. We should at least be asking why we haven't.

So what can you actually do this week? If there's a teenager or a partner in your life using a companion app, just ask them about it. Not as an interrogation, as a real conversation. Ask what the AI remembers about them. Ask how it makes them feel when they're not using it. Common Sense Media keeps independent reviews and age guidance on AI apps, including companion apps, and it's a genuinely useful place to start before you hand a device to a kid or worry about one your teen already has. You don't need a law to start paying attention. You just need to start.