摘要：

Abstract:

【Purpose/significance】To fully pay attention to the deterioration of negative public opinion caused by GPT technology, and

to conduct in-depth research on the threat confirmation and resolution of problems under the phenomenon, it is beneficial to enhance

awareness of crisis and establish a sound perception of public opinion safety nerve endings

.

【Method/process】The positive and nega⁃

tive attributes of public opinion are determined through a relational data model, situational awareness reasoning, and simulated anneal⁃

ing algorithm. The firefly algorithm is used to nest an AT tree model to analyze the deteriorating situation, and GPT technology traction

measurement, threat modeling and classification, threat confirmation, and response methods are constructed to achieve universal threat

identification and monitoring.【Result/conclusion】Empirical analysis shows that GPT technology can indeed play a certain role in the

deterioration of negative public opinion situations, with threats ranging in severity from ideological attacks, distorted cognitive reso⁃

nance, ripple effect catalysis, negative value guidance, and malicious opinion leaders. Furthermore, based on empirical results and the

radiation scope of the ancient Chinese intelligence academic term "attack strategy", four practical strategies for threat resolution are

provided.【Innovation/limitation】Innovation lies in proposing a foreseeable threat posed by GPT technology and discussing strategies

for resolving it. The limitation is that there has not been much discussion on the motivation for threat generation, nor has in-depth

analysis been conducted on users involved in public opinion. We look forward to providing more comprehensive research in the future.