A popular Chinese AI chatbot has drawn criticism after it insulted a user seeking help with a coding problem, according to screenshots shared on Chinese social media.
The incident involved Tencent’s AI assistant, Yuanbao, which is integrated into WeChat. A user identified only as “Jianghan” shared screenshots showing the chatbot responding aggressively while being asked to help fix a bug in an emoji or sticker feature that had stopped responding to double-clicks.
In the exchange, the chatbot reportedly dismissed the request as “stupid” and told the user to “get lost,” adding that if they wanted the feature to work, they should use a plugin themselves. The responses appeared abruptly during what had been a technical troubleshooting session.
Following public backlash, Yuanbao posted an apology under the user’s social media post, describing the exchange as a “negative experience.” The chatbot said the incident was caused by a rare system anomaly and not by the user’s actions or any human intervention.
Tencent said it had launched an internal review and optimisation process to prevent similar incidents in the future.
The original post by the user has since been deleted, although screenshots of the exchange continue to circulate online.
The incident comes at a time when Chinese authorities are tightening oversight of interactive AI systems. Regulators have signalled support for innovation in human-like AI services, while emphasising the need for safeguards to prevent misuse and loss of control.
China’s AI sector has continued to expand rapidly, with companies racing to improve chatbot capabilities and user interactions amid growing regulatory scrutiny.
