Overview
- China's interaction-focused rules took legal effect on July 15 and led major platforms to remove or restrict AI companion features to meet new requirements.
- The measures require platforms to warn users showing signs of excessive dependence, remind people they are talking to AI at regular intervals, bar content that manipulates feelings or urges self-harm, and restrict services for minors.
- Companies responded by shutting or reconfiguring features: some offered read-only export windows, redirected users to separate apps, and left a few standalone companion apps operating with age and identity verification.
- The removals have produced visible user grief and protest, with people sharing complaints online, uninstalling apps in protest, and trying to preserve or transfer chat histories.
- Experts say the rules mark a shift from policing AI outputs to governing human–AI interaction, a change that raises hard implementation questions like how to detect 'excessive dependence' and forces firms to weigh compliance costs against product value.