What Is Cyberbullying?

Cyberbullying is the use of digital communication technologies to repeatedly harm, harass, threaten, or humiliate an individual. Unlike face-to-face bullying, cyberbullying can occur continuously through social media platforms, messaging applications, online gaming environments, and email, reaching targets at any time and in any location. The persistent and often anonymous nature of online harassment distinguishes it from offline counterparts and can amplify its psychological impact.

The phenomenon emerged alongside the proliferation of mobile devices and social networks in the early 2000s. Research across psychology, communications, and computer science has since examined its prevalence, mechanisms, and the technical means available for detection and mitigation.

Forms and Mechanisms

Cyberbullying takes several documented forms. Direct harassment involves sending threatening or abusive messages to a target. Cyberstalking refers to sustained monitoring and intimidation through digital channels. Doxing, the public disclosure of private personal information, is used to expose targets to broader harassment. Impersonation involves creating fake accounts or messages to embarrass or deceive a victim. Exclusion from online communities and the non-consensual sharing of intimate images represent additional categories that carry significant legal consequences in many jurisdictions.

Research published in the Journal of Cybersecurity on cyberbullying definitions and prevalence documents the definitional challenges in cross-disciplinary work, noting that inconsistent definitions across studies complicate measurement of true prevalence. A recurring finding is that cyberbullying disproportionately affects adolescents, with studies drawing on NIH data indicating elevated rates of anxiety and depression among those targeted online.

Detection and Computational Approaches

The scale of online interaction has made manual content moderation insufficient, driving research into automated detection methods. Natural language processing and machine learning classifiers trained on labeled datasets are now widely used to flag abusive content. Graph-based methods analyze interaction patterns among users rather than message content alone, capturing coordinated harassment campaigns that individual message screening misses.

IEEE research on cyberbullying and cyberviolence detection frames the problem as a triangular analysis of users, their activity patterns, and message content, finding that combining all three signal types improves detection accuracy over single-feature approaches. Platforms deploying such systems face a tension between effective moderation and preserving free expression, a challenge that has drawn increasing regulatory attention in multiple jurisdictions.

Impact and Response

The psychological consequences of cyberbullying are well-documented. Victims report higher rates of depression, social withdrawal, academic difficulties, and in severe cases, self-harm. Studies reviewed by the NIH on cyberbullying and adolescents confirm a link between online victimization and adverse mental health outcomes, while also noting that individual resilience factors, social support, and parental involvement moderate these effects.

Legislative responses have expanded considerably. Many US states and a growing number of countries have enacted laws specific to cyberbullying, adding to existing harassment and stalking statutes. Unsolicited bulk messaging, or spam, overlaps with cyberbullying in cases of mass targeted harassment campaigns, and anti-spam technical standards have been adapted to address coordinated abuse.

Applications

Cyberbullying research and detection technology has applications in a wide range of areas, including:

• Social media content moderation and trust-and-safety systems
• School and campus anti-harassment programs
• Mental health intervention and youth protection services
• Law enforcement digital evidence analysis
• Platform policy design and regulatory compliance