Publications

A report by Vys and Social & Media Matters documenting anonymized insights from a closed-door multi-stakeholder roundtable convened in New Delhi on Oct 6, 2025, ahead of India’s inaugural Trust & Safety Festival. Bringing together Indian youth, civil society leaders, researchers, leading platform representatives, and expert policy practitioners, the discussion examined how AI systems, platform governance, and product design choices intersect with the linguistic diversity across Indian youth, the social hierarchies they must navigate, and the rapidly evolving digital landscape they experience. The report distills key themes, tensions, and areas of consensus, and offers concrete product, policy, and ecosystem-level recommendations to strengthen youth safety, trust, and accountability in the Indian context.

A report by Resolver, with contributions from Vys, examining the complex and persistent ecosystem of online harm commonly referred to as “the Com.” Drawing on multiple years of intelligence and investigative research, the report maps how fragmented online subcultures and overlapping behaviors can evolve and escalate into coordinated harm targeting children and young people. It outlines the structural drivers behind these dynamics and offers practical implications for regulators, platforms, and those responsible for governance, trust, and safety.

A report by Vys advisor Vaishnavi J in The ANNALS of the American Academy of Political and Social Science examining the persistent gap between external platform governance efforts and how product decisions are actually made inside technology companies. The article unpacks the internal dynamics of design, experimentation, and iteration, and proposes a framework for reconnecting academic and regulatory objectives with the operational centers of platform power. It offers a practical lens on how governance can move from theory into real-world product decision-making.

A code developed by 5Rights Foundation, with contributions from Vys, setting out a practical framework to identify, assess, and mitigate the risks AI systems pose to children. The Children & AI Design Code provides lifecycle-wide guidance for those who build and deploy AI, requiring foreseeable risks to children to be addressed by design and by default. Spanning contextual analysis, risk criteria, development-stage considerations, and a clear implementation checklist, the Code translates children’s rights into concrete product, governance, and engineering actions to ensure AI systems better support children’s safety and development.

A report by the Institute for Security and Technology, supported by Vys, examining how generative AI — particularly social conversational agents — may reshape social cohesion. Produced through the Generative Identity Initiative, this year-long, multi-stakeholder collaboration explores how GenAI can influence cognition, interpersonal trust, wisdom, and collective memory. The report outlines core areas of concern and presents a comprehensive 27-point research agenda to guide industry, academic, and civil society efforts in understanding and mitigating emerging risks.