Digital platforms have become central to social interaction, identity formation and community participation, but their primary design objectives — play and competition for games, expression and audience-building for creator platforms, communication and discovery for social media — do not always align with community wellbeing, according to an editorial published in Frontiers in Public Health on 20 November 2026.

The editorial, written by George Ng and Constance Wang of the University of California, Berkeley's School of Public Health, introduces a Research Topic collection of studies examining the behavioural and mental health impacts of negative online interactions. It argues that content moderation, while visible and measurable, is an incomplete response to the harms that can arise in digital spaces.

Under the European Union's Digital Services Act (DSA), very large platforms and search engines must assess and mitigate systemic risks from their design and algorithms, including risks to mental wellbeing. But the editorial notes that observing moderation actions — such as the number of removals or restrictions — does not establish their effects on psychological health. An increase in removals could reflect better detection, more harmful behaviour, or changes in enforcement policy, the authors write.

The collection of studies included in the Research Topic illustrates why a broader perspective is needed. A review by Li of social media use among college students describes opportunities for belonging and support alongside associations with anxiety, depression and problematic use. Keum et al. found that online mental health engagement strengthened associations between online racism and certain adverse outcomes among Asian American and Pacific Islander emerging adults, with a narrower pattern among Black emerging adults. The editorial notes that these cross-sectional results cannot establish causality, but they show how ostensibly supportive activities may coexist with exposure to hostility.

Culture also shapes both digital experiences and the research questions asked about them. Cook et al. identified an emphasis on addiction and adverse outcomes in the East Asian literature they reviewed, with relatively limited attention to technology-enabled wellbeing. Together, the papers caution against treating online participation as uniformly beneficial or harmful, or using engagement as a proxy for health, according to the editorial.

The negative impacts of online behavior ripple beyond one-on-one exchanges. Using an ecological model, Marshall and colleagues map how digital hostility affects pupils, households, teachers, and schools, suggesting interventions at multiple tiers. In her study of gaming spaces, Kowert points to clear, context-appropriate behavioral guidelines and active community involvement as additions to punitive measures.. The editorial notes that these approaches are promising but that evidence on their long-term efficacy and generalisability remains limited, and their value should be investigated rather than presumed.

An ecological perspective also complicates responsibility, the authors argue. Platforms design architectures for discovery, interaction, incentives and enforcement, but users produce much of the content and behaviour that makes communities valuable and sometimes harmful. Community leaders influence norms; schools, families and other institutions can offer support; policymakers establish wider expectations and accountability. Responsibilities are therefore shared, but not interchangeable, the editorial states.

"Platforms cannot directly control every interaction they host, yet their design and governance shape which behaviors are enabled, amplified, or constrained," the authors write. They add that shared responsibility should not shift the burden of preventing harassment onto those subjected to it.

The editorial also notes that safety, inclusion and community health are not identical goals. Preventing harassment or discriminatory exclusion differs from actively fostering belonging. Not every digital space needs to serve every audience or offer the same social experience: competition, provocative art and specialised interests can be legitimate features. Nevertheless, choices that benefit one group or advance a product objective may impose costs on others. A design that increases discovery or interaction may also increase unwanted exposure, while stronger participation controls could limit opportunities for connection. The authors describe these as potential tradeoffs to investigate rather than universal effects to assume.

Evaluating these tradeoffs is difficult, the editorial acknowledges. Design and governance interventions influence interactions indirectly, while users select into communities and change behaviour over time. An apparent improvement may reflect reduced harm or the departure of those most affected. Fewer reports can signal greater safety or diminished trust; higher engagement can reflect connection or conflict. Even a successful intervention may alter reporting behaviour or community composition, complicating efforts to determine its actual effects. Moderation accuracy, enforcement consistency and complaint volumes are useful operational indicators, the authors write, but not substitutes for psychological outcomes.

The editorial calls for future research to connect governance and design interventions to repeated victimisation, perceived safety, belonging, participation and psychological distress. It recommends longitudinal research, natural experiments, controlled evaluations and cross-cultural replication to help distinguish causal effects from selection and changing community composition. Partnerships between researchers, platforms and affected communities may improve measurement while requiring careful attention to privacy and research ethics.

Rather than seeking a universal community-health score, the authors argue, research should ask which interventions improve which outcomes, for whom, and under what conditions.

"The articles in this collection do not establish a single model for healthy digital communities," the editorial concludes. "They instead point toward a more demanding public health agenda: understanding how platforms, participants, institutions, and policymakers can reduce avoidable harm while sustaining the diverse forms of participation that give digital communities their value."

Ng disclosed that he is Co-Founder and Chief Executive Officer of GGWP, a company providing AI-based content moderation and community safety technologies, and holds an equity interest in the company. Wang declared no competing interests.