China's AI companion rules leave enterprise agents untouched
China's AI companion rules hit consumer apps hardest, leaving most channel-built enterprise and productivity agents outside the new regime.
China has written the world's first national rulebook for AI companions, the chatbots people build to talk to like a friend or partner rather than to get work done. As those rules near their July 15 start date, the country's biggest consumer AI apps, ByteDance's Doubao, Alibaba's Qwen and Tencent's Yuanbao, have switched off the features that let users create their own bots.
The good news for the channel is that the rules were never aimed at them. What Beijing wants to police is the consumer companion, not the service desks and workflow copilots that partners deploy for their clients.
The retreat has been quick. Doubao told users its agent function would go offline on July 15, blaming "product function adjustments." Qwen's humanlike and user-created agents stopped on July 10, with its wider agent services following five days later. Tencent had already pulled a comparable Yuanbao feature in June. Millions of users are affected, and many have complained on Weibo about losing chat histories they had built up over months.
The regulation driving all this is the Interim Measures for the Administration of AI Anthropomorphic Interactive Services, issued on April 10 by the Cyberspace Administration of China, alongside four other agencies, including the Ministry of Industry and Information Technology and the State Administration for Market Regulation.
It applies to services that mimic a real person's personality, thinking and way of speaking to keep up an ongoing emotional relationship with the user.
Where the channel sits
That definition is the whole story for partners. Customer service bots, knowledge Q&A tools, workplace assistants and education software are carved out, as long as they steer clear of that sustained emotional engagement. Most enterprise agent work, from service-desk automation to internal copilots, lands on the safe side of the line.
Safe does not mean unaffected, though. Any provider that adds anthropomorphic features, or grows past one million registered users or 100,000 monthly activities, has to pass a security assessment covering eight areas and file its algorithms with provincial regulators. App stores must check that this has been done and pull anything that hasn't.
The rules carry no explicit reach beyond China's borders, but the CAC has a track record of enforcing its data and cybersecurity laws against overseas firms with Chinese users, so a foreign vendor is not automatically out of scope. The measures are also silent on who takes the blame when a breach comes from the underlying model rather than the platform running it, which is worth raising with any client building on someone else's model.
The one thing to watch is that phrase, emotional interaction. A customer-facing bot designed to build rapport could drift from an exempt tool into a regulated companion without anyone intending it to.
Rebuilding, not retreating
None of this reads as a vendor pulling back from agents. ByteDance is already steering Doubao users to Maoxiang, a separate app built to run agents under the new rules. Alibaba has offered Qwen users no such route, though neither company has signalled any let-up in agent spending, and Gartner expects up to 40% of enterprise apps to carry task-specific agents by 2026.
So far, enforcement seems to be running ahead of the deadline. Shanghai's internet regulator said on June 26 it had stripped more than 14,000 non-compliant agents from platforms, pointing to bots impersonating officials, vulgar role-play and the quiet harvesting of personal data.
The message to anyone placing AI bets in China is plain enough: Beijing will back agents as serious business infrastructure, and it will lean hard on the consumer-companion edge. The enterprise lane is open. The compliance load that comes with scale is now part of the job.