Intervention analysis for fake news diffusion: an evolutionary game theory perspective

Controlling the dissemination of fake news is paramount for societal well-being. This paper explores how to govern fake news on social media, considering the reputation of both government and social media platform, reward and punishment mechanisms, as well as the associated costs and benefits from the perspective of evolutionary game theory. Within the game, the government can opt for supervision or no supervision, the social media platform can decide whether to respond to government strategies actively or not, while individuals can choose to forward or abstrain from forwarding fake news. The results show that the system will stabilize in three states contingent upon initial parameters: (not forward, not active, supervise), (forward, not active, supervise), and (forward, active, supervise). Key factors influencing the governance of fake news are also identified. Individually, reducing the benefits of forwarding fake news and increasing penalties from both the government and social media can discourage individuals from sharing fake news. On the social media level, minimizing the cost of their responses to government regulation, decreasing the revenue they gain from individual retweeting of fake news, and increasing the benefits of their responses to government policies can incentivize proactive social media responses. At the government level, enhancing rewards for social media compliance with government policies and enforcing penalties for non-compliance can motivate social media to align with regulatory strategies. The findings in this paper provide valuable insights into governing fake news.