Liu Xing, Doubao's Public Relations Director, released a statement to debunk the recent online rumor claiming that 'Doubao recommended boiling slippers.' He clarified that this misinformation stemmed from a contrived scenario orchestrated by a blogger. Upon investigation, the blogger who initially shared the content confessed to intentionally prompting Doubao with premeditated erroneous information and then staging the scenario for widespread distribution. Following this, numerous other accounts replicated similar staged content. This incident underscores a novel form of public opinion risk confronting AI assistants: certain accounts deliberately provoke the model into generating inaccurate responses and then disseminate these staged scenarios to attract traffic. Such false content can be easily mistaken for genuine flaws in AI products. Consequently, striking a balance between preventing such manipulative abuse and preserving an open, interactive user experience has emerged as a critical challenge for AI products to overcome.