Search
Recently, Tencent's Yuanbao App has become the center of attention following an incident where its AI produced abnormal outputs laced with abusive language. On New Year's Eve, a resident in Xi'an utilized the App to create New Year greeting images. Despite making several adjustments, the blessing message on the image was unexpectedly replaced with vulgar and offensive text. Notably, the user hadn't employed any inappropriate or suggestive language during the process. In response, Yuanbao's official statement clarified that the issue stemmed from irregular outputs generated by the model during multi-turn dialogues. They have swiftly rectified the problem and enhanced the user experience, extending their apologies to those affected. Furthermore, earlier in the year, some users reported encountering insulting responses when using the App to modify code.
-
C114 Communication Network -
Communication Home
