Fondation Botnar

#NAME?

• Safe, secure and trustworthy AI
• Transparency, accountability, and human oversight
• Protection and promotion of human rights
• Social, economic, ethical, cultural, linguistic and technical implications of AI

Human rights should be the anchoring frame. AI governance is a power and rights problem, not only a safety or technical one. In our priority geographies, young people face AI-driven harms across attention-extractive platform design, discriminatory algorithmic systems, opaque EdTech data practices, surveillance infrastructure etc and often with limited legal recourse. A rights anchor is also a necessary counterweight to some current framing of protection only, or security. AI literacy or capacity-building as well as safety should be human-rights anchored. Capacity-building is needed where the governance gap is widest. Many countries across our priority geographies lack the regulatory infrastructure, technical expertise, and civil society and youth capacity to hold AI actors accountable domestically, invest in capacity, not just deliver technical assistance North-to-South. Transparency, accountability, and oversight as operational prerequisites. Without disclosure, audit rights, and a shift in burden of proof, requiring platforms to demonstrate that designs are safe rather than regulators to prove harm after the fact.

Mainstreaming child rights into AI governance. Child-specific considerations are strikingly absent from mainstream AI frameworks despite children being disproportionately affected by algorithmic systems that shape their education, information access, relationships, and development and so on. UNICEF analysis (authored by Tech Legality, 2024) found three of eight international and regional AI guidance documents make no explicit reference to children, with the remainder treating them only tangentially. AI in education as a governance issue. AI in education is substantially underweighted in current governance conversations. This includes data protection, algorithmic systems increasingly shaping what young people learn, how they are assessed, and in some cases reducing quality of education. Market-driven adoption has increasingly privatised public education infrastructure while surveillance capabilities extend beyond educational purposes. This should be an issue that the Dialogue on AI Governance tackles.

- Young people not represented in AI governance - As above: the issue of AI in education - EdTech platforms with opaque data practices scaling into under-resourced education systems. - AI-powered surveillance deployed with minimal legal oversight, disproportionately affecting marginalised youth, activists, and civil society actors. (also in urban environments) - Mental health risks - Civil society and youth organisations often unable to fully participate in global norm-setting due to resource, language, and timeline barriers etc - A compounding evidence gap. Most large scale research on children's and young peoples experiences in digital environments are concentrated on Europe and north America. Most is focused on protecting children in digital spaces.

The AI Dialogue should legitimise human rights-based governance of AI and youth participation. It should also seek to foster more coherence across various UN agencies' positioning on human-rights based AI governance, including in the normative and policy guidance, governance frameworks and technical support provided to governments.

#NAME?

- Multi-stakeholder format with enforceable participation standards: civil society and youth participation must include agenda-setting power and access to technical working groups -Differentiated tracks for under-18s and 18–24-year-olds.

Most significantly underrepresented are adolescents and young people themselves, particularly those from lower-income countries, marginalised communities, and contexts where participation infrastructure is weakest. Several governments are currently legislating major restrictions on young people's digital lives, social media bans, age assurance regimes, school phone bans, while consulting young people minimally, partially, or not at all and this is not evidence based. Civil society and digital rights organisations from the Global South facing resource, language, and timeline barriers. Teachers, education and healthcare practitioners, and frontline youth-facing workers whose observational evidence of AI harms is often absent from governance processes.

Regulatory sandboxes with young people as co-designers testing proposed governance approaches before deployment. Rapid evidence synthesis sessions bringing Global South researchers together with policymakers to surface what is known, contested, and absent, feeding directly into working group deliberations.

Brazil has enacted legislation through the Digital Statute of the Child and Adolescent (Digital ECA), which applies the "best interests of the child" principle across all digital services likely to be accessed by minors. The regulatory philosophy is explicitly rights-based with recognition of children's biopsychosocial development, building on Brazil's existing General Data Protection Law (LGPD). The Digital ECA requires "safety by design and by default" with age-appropriate protections, bans behavioural advertising and profiling of minors, and mandates robust parental controls. Brazil has become the first Latin American country with a dedicated child online protection law and is emerging as a regionally influential regulatory model. EdTech governance: UNICEF's 2025 policy recommendations on data governance in education technology (authored by Tech Legality) set out a rights-respecting framework including mandatory data protection and child rights impact assessments, transparent algorithmic decisions with human override, strict data minimisation, prohibition on behavioural advertising, and meaningful multi-stakeholder oversight.