Collaborative Governance and Decision Support Strategies for Misinformation Generated by Generative Artificial Intelligence

The rapid rise of generative artificial intelligence (GAI) has intensified misinformation creation and dissemination, disrupting information order and posing major challenges to governance. Misinformation not only shapes public opinion but also threatens political, economic, and social stability, highlighting the need for effective collaborative governance and decision support strategies. This study introduces a fusion framework using Self-Attention Generative Adversarial Networks (SAGAN), Variational Autoencoders (VAEs), and Transformers for misinformation detection. Evaluated on four public datasets, including News Media Bias and Misinformation Detection Dataset (NELA-GT), the method outperforms traditional algorithms in accuracy, efficiency, and stability. On NELA-GT, it reduces FLOPs by 46.6%, inference time by 48.3%, and improves Mean Absolute Percentage Error (MAPE) by 38%. The proposed algorithm significantly enhances detection speed and predictive capability, showing clear advantages in complex scenarios, and offers strategic guidance for preventing, detecting, and managing misinformation generated by GAI.