Abstract
In recent years, alongside the steady growth of China's internet user base, diverse forms of cyber violence—including doxxing, rumor-mongering, defamation, and the malicious spread of manipulated content—have emerged as frequent, widespread, and highly contentious issues. These phenomena pose significant potential harm to individuals and society. While the Chinese government and relevant authorities have prioritized this problem, enacting a series of policies and regulations, the inherent ambiguity and evolving nature of cyber violence present considerable challenges for both research and governance. This study, therefore, employs empirical investigation to further analyze the current state of cyber violence and its influencing factors. Based on the findings, the research proposes a multi-level strategy to combat cyber violence, aiming to contribute to building a safer digital space.
Introduction
As of mid-2025, China’s internet user base has surpassed 1.1 billion individuals. This unprecedented digital expansion has outpaced the evolution of corresponding legal and social frameworks, giving rise to "online disorders" such as doxxing, coordinated defamation, and the spread of misinformation. Cyber violence is distinct from physical violence due to the "traffic-driven" incentives of the digital economy and the psychological shielding provided by online anonymity.
A significant challenge in current governance is the "regulatory grey area," where behaviors like "indirect" verbal attacks or the public exposure of private information occupy a space between social impropriety and clear illegality. This study seeks to understand the psychological and environmental drivers that transform netizens into perpetrators to inform more effective intervention strategies.
Methodology
The researchers employed a mixed-methods approach to analyze the dynamics of cyber violence:
- Theoretical Framework: The study is grounded in Albert Bandura’s Triadic Reciprocal Determinism and Cognitive Behavioral Theory, analyzing the bidirectional interactions between internal cognition, behavior, and the environment.
- Empirical Investigation: Surveys and in-depth interviews were conducted across three representative cities in Zhejiang Province: Wenzhou, Taizhou, and Lishui.
- Variable Analysis: The team examined five independent variables—attitudes, subjective cognition, social environment, online norms, and "propensity for cyber violence"—to determine their impact on abusive behavior.
Key Findings and Demographic Insights
The empirical data revealed several critical insights into the nature of online aggression:
- The "Susceptible Profile": The highest risk for engaging in cyber violence was found among males, students, and young adults (ages 25–30) in early career stages. These groups often experience high life pressures and may lack adequate coping mechanisms, leading to online venting.
- The Definition Gap: While explicit insults are widely recognized as violence, there is significant divergence regarding doxxing and "implicit" profanity. This ambiguity contributes to cognitive biases that can lead to accidental participation in harassment.
- Internal vs. External Drivers: Subjective cognition—specifically how powerful or anonymous a user feels—was identified as a stronger predictor of violent behavior than external social environments or online norms. Factors influence engagement in the following descending order: Subjective Cognition > Social Environment > Online Norms.
- The Bystander Paradox: Netizens often remain indifferent as bystanders but become highly proactive in seeking legal recourse when they are victims, suggesting that personal self-interest currently outweighs social morality as a deterrent.
A Multi-Level Strategic Framework
To address these challenges, the study proposes a comprehensive "Multi-Level Approach." This strategy moves beyond reactive measures to create a synergistic ecosystem of governance.
The framework integrates three primary pillars:
- Government Regulation: Establishing clearer legal boundaries to eliminate "regulatory grey areas" and providing a robust framework for enforcement.
- Platform Moderation: Leveraging advanced technologies, including Large Language Models (LLMs) and AI, for real-time monitoring, automatic classification, and accurate labeling of harmful content to move beyond manual, reactive moderation.
- User Digital Civility: Enhancing digital literacy and social morality to reduce the "propensity for cyber violence" at the individual level.
Conclusion and Future Directions
While this study provides a foundation for multi-level governance, it is limited by its geographic focus on Zhejiang Province and the potential for self-reporting bias. Future research should expand the demographic scope and differentiate between various types of cyber violence (e.g., political vs. personal) to evaluate the effectiveness of specific interventions. By fostering collaboration between the state, technology providers, and the public, it is possible to mitigate the harms of cyber violence and build a more resilient digital space.
Summary updated