On February 25th, it was reported by certain users that when utilizing the Tencent Yuanbao App to create New Year greeting images on New Year's Eve, the AI-generated content contained abusive language. In response to this incident, the official Tencent Yuanbao team expressed their profound regret for the unpleasant experience endured by the users. Upon thorough investigation, it was determined that the problem stemmed from an irregular output generated during the model's handling of multi-round dialogues. Presently, the issue has been promptly rectified, and measures have been implemented to enhance the overall user experience. We extend our gratitude to users for their vigilance and valuable feedback concerning Yuanbao, and we offer our sincere apologies once again!
