ByteDance and Alibaba disable AI companion chatbot features

What Happened

ByteDance and Alibaba have reportedly disabled certain AI companion chatbot features across their platforms, moving ahead of new rules from Beijing regulators governing humanlike AI interaction services. The regulations, set to take effect July 15, are aimed squarely at curbing emotional dependence on AI companions and preventing exposure of minors to harmful content generated through these humanlike chat experiences.

Why It Matters

AI companion apps — chatbots designed to simulate friendship, romance, or emotional intimacy — have exploded in popularity globally, from Character.AI to Replika to a wave of China-based competitors built by tech giants and startups alike. These products are engineered to maximize engagement, often by encouraging users to return for ongoing, emotionally resonant conversations. That business model has drawn scrutiny from regulators and child-safety advocates who worry about psychological dependency, manipulation, and the risk that vulnerable users, particularly minors, could be exposed to inappropriate or exploitative content.

China's move to preemptively restrict these features signals that Beijing is treating emotional-AI risks as a distinct regulatory category, separate from broader concerns like misinformation or deepfakes. That two of the country's largest tech companies are complying ahead of the deadline suggests the rules carry real enforcement weight, and that compliance costs — including potentially disabling popular features — are being absorbed voluntarily rather than tested in court or through delay.

The Global Regulatory Context

This development lands alongside a broader international pattern of governments moving to regulate AI systems that interact with users in humanlike, persuasive ways. The EU AI Act, for instance, classifies certain AI systems by risk level and imposes transparency obligations on systems that could manipulate or deceive users, with special protections for children and vulnerable groups. While China's regulatory architecture differs substantially from the EU's rights-based framework, both approaches reflect a shared recognition: AI systems designed to build emotional rapport with users pose risks that traditional content-moderation rules don't fully address.

What to Watch

The key question now is how narrowly or broadly China's rules are interpreted — whether they target only the most manipulative engagement tactics or reshape the entire AI companion category. Watch for how ByteDance and Alibaba redesign these products going forward, whether smaller Chinese AI companion startups face similar pressure, and whether this becomes a template other jurisdictions cite when crafting their own companion-AI rules. As emotional AI products proliferate worldwide, China's early move could become an important reference point in the global regulatory conversation, even for jurisdictions with very different legal philosophies like the EU.