UNICEF

For the first Global Dialogue to be a success, UNICEF recommends that discussions be structured around specific questions and produce concrete outputs, including: priority areas for follow-up; proposed areas of UN system coordination; draft principles for inclusion in future instruments; and practical models and tools for rights-based AI governance. Session chairs should be tasked with synthesising the discussions into actionable summaries for the co-Chairs. Specifically, UNICEF recommends the Global Dialogue aim to produce the following outcomes: • A plan to coordinate UN system actions on AI governance and translate existing commitments into practical tools for governments and businesses • A Co-Chairs' Summary explicitly addressing children's rights as a cross-cutting and priority issue, and the participation gap for children and youth in AI governance, as well as concrete follow-up actions • Agreed principles or recommendations on the inclusion of children's rights standards in national and international AI governance, building on existing international human rights law frameworks and global standards on responsible business conduct • A roadmap for meaningful integration of children and youth participation into ongoing Global Dialogue processes, including the preparatory architecture for future sessions • Establishment and strengthening of mechanisms for sustained, coordinated attention to children's rights and interests within the international AI governance architecture. This should include identification of existing mechanisms and platforms, including UNICEF-led or co-led initiatives, that can serve as vehicles for implementation between Dialogue sessions • Translation of research findings by the Independent International Scientific Panel into accessible, recommendations relevant for policymakers and industry, including on the state of evidence regarding AI's impacts on children and youth.

• Safe, secure and trustworthy AI
• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights

UNICEF's selection reflects its mandate to uphold the rights of every child in an AI-shaped world, and prioritises the thematic areas most directly implicated in children's present and future well-being - while emphasising the critical importance and relevance of all themes listed . Protection and promotion of human rights is UNICEF's primary lens. Children are rights-holders under the Convention on the Rights of the Child, yet child rights protections remain absent from most AI governance instruments. The Dialogue must address this gap, including by clarifying how international human rights law applies to AI systems affecting children. Safe, secure and trustworthy AI is of concern for all children, recognising they are not a homogenous group and that certain groups of children have unique needs that must be addressed and may face distinct and disproportionate risks online - including those facing discrimination based on grounds of sex, disability, race, ethnicity, religion, migration status, indigenous status, or other status as recognised under international human rights law . AI that lacks safety-by-design, explainability and appropriate safeguards risks exposing children to distinct harms, from algorithmic manipulation and privacy violations, to content (e.g., AI-generated child sexual abuse materials) that endangers their safety and wellbeing. Social, economic, ethical, cultural, linguistic and technical implications of AI directly shape children's development and future. The effects of AI on cognition, identity formation, learning and social behaviour are significant and still emerging. Decisions made today will determine the world that children and young people inherit, making it essential that they are visible and heard in these debates. AI capacity-building is essential to close the access and opportunity gap for children, especially in low- and middle-income countries. Equitable AI governance must include targeted investment in digital literacy, connectivity, and locally relevant AI tools, ensuring AI's benefits reach every child, including the most marginalised.

Children's rights, needs, safety, and unique stages of development are directly implicated across all of the themes listed, and encapsulate the full spectrum of AI governance issues, covering both protection from risks and harms, as well as equal access to opportunities - including for example opportunities to gain digital skills and AI literacy, and the responsible use of AI to achieve positive outcomes for children and advance progress towards the SDGs in areas such as education, health, gender equality, water, sanitation and hygiene, emergency response and humanitarian action, social inclusion and child protection. Children are among the fastest-growing users of AI, and often the most affected. Yet children are not a homogenous group - different children experience AI's impacts in distinct ways, shaped by age, gender, disability, ethnicity, migration status and other factors. UNICEF recommends that each thematic cluster explicitly name children and youth as a priority and include a specific lens on these differential impacts. The impacts of AI on girls and adolescent girls warrant specific attention, with girls facing disproportionate exposure to AI-enabled harms, including non-consensual intimate imagery and deepfakes, online sexual exploitation, algorithmic surveillance and stalking, and gendered bias embedded in AI models that can shape their access to education, information and services. AI governance frameworks should therefore reflect a gender-responsive child rights approach, attentive to both the heightened risks that girls face and the equity opportunities AI can unlock when designed with and for them. In addition to mainstreaming children's rights, needs, safety and development across all tracks, UNICEF recommends establishment of a dedicated thematic cluster on Children, Youth and AI. UNICEF welcomes that the thematic discussions will be co-chaired by a Member State and a relevant stakeholder, and stands ready to co-facilitate this work, building on its mandate, expertise and extensive field presence.

AI is reshaping the lives of children and the world they will inherit at an extraordinary pace, transforming how they learn, play, develop, communicate, and how they access information, health services, and social protection. It will also radically transform the world's job markets and revolutionise the type of competencies and skills children will need to access employment and thrive as future adults. Moreover, children's rights and well-being are impacted by the use of AI around them – such as in education and social services. Children are among the fastest-growing users of AI, and face distinct opportunities and risks that the current global AI governance frameworks do not yet adequately address. Children will be the primary inheritors of the AI systems being built today, so it is important to also include a focus on potential future impacts of AI policies on rights of future generations of children in a world that will be shaped by today's AI decisions. At the same time, there is limited evidence available on how AI technologies are being used by, and for, children - the data concerning AI's effects on children, including on cognitive and emotional development, identity formation, and social behaviour, is still emerging. This limits the evidence base for proportionate and effective policy responses. But children's exposure to AI means their rights, safety and unique needs should be explicitly prioritised, protected and respected in AI design, development and deployment contexts. Child rights should be expressly integrated into AI governance frameworks and approaches to responsible business conduct including risk assessment practices and embedding children's rights and safety by design, rather than attempting to retrofit protections after harms have already occurred. UNICEF highlights the following critical gaps in current AI governance: i. Evidence gap - insufficient research and data on the impacts of AI on children's cognitive, social, emotional and physical development, particularly as these impacts unfold over time and across diverse contexts. ii. Participation gap - absence of meaningful, safe and inclusive mechanisms for children and youth to contribute their voices, experiences and perspectives to AI policy and governance processes at national and international levels. iii. Governance and institutional coordination gap - absence of dedicated, coordinated and sustained attention to children's rights and interests across global AI governance mechanisms iv. Access, equity and opportunity gap - persistent disparities in children's access to AI technologies, digital skills and the capacity needed to benefit from and participate in AI systems, with particular concern for children in low- and middle-income countries, or in marginalised communities and for girls, who continue to face a persistent gender digital divide in connectivity, device access and AI/STEM skills.. v. Industry implementation gap - failure to systematically embed children's rights and child safety principles into the design, development and deployment of AI systems.

UNICEF considers the primary purpose of the Global Dialogue on AI Governance to be the establishment of a legitimate, inclusive and action-oriented multistakeholder forum that can give political direction to emerging international norms on AI governance, while ensuring meaningful participation of Governments, civil society, the private sector, academia and international organisations. From a UNICEF perspective, the Dialogue's most important role and added value in advancing international cooperation on AI governance lies in its potential to: • Elevate under-represented interests (including those of children, youth, and populations in low- and middle-income countries) within global AI governance debates – including also attention to girls and adolescent girls who are often unheard and unseen • Bridge the gap between AI technical and standards communities (where children's rights considerations have been largely absent) and intergovernmental norm-setting processes of the UN system • Reaffirm the role of responsible business conduct to ensure respect for children's rights in the digital environment, and clarify the application of the UN Guiding Principles on Business and Human Rights in the context of AI development and deployment • Create a platform to translate international children's rights frameworks (including the Convention on the Rights of the Child) into concrete, agile and operational AI governance tools and standards • Focus on practical solutions, tools and case studies that can be implemented at national and regional levels.

The Global Dialogue should aim to build on, and connect to, a robust body of existing work to ensure coherence, avoid duplication, and amplify proven AI governance frameworks. The distinctive added value of the Global Dialogue lies in its potential to convene these parallel workstreams into a coherent, authoritative and action-oriented framework, establishing shared principles, surfacing governance gaps, and creating co-ordination mechanisms that no single body or initiative has yet achieved. Within the UN system, the Global Dialogue should connect with: the Global Digital Compact and its implementation architecture; the work of the UN High-level Advisory Body on AI and its successor mechanisms; the Convention on the Rights of the Child and authoritative interpretations by its treaty body including General Comment 25 on children's rights in the digital environment and General Comment 16 on State obligations regarding the impact of the business sector on children's rights; the UN Guiding Principles on Business and Human Rights; the UN System Joint Statement on AI and the Rights of the Child; UNESCO's Recommendation on the Ethics of AI; Across the regional AI governance ecosystem, the Global Dialogue should engage with: the OECD AI Principles; the Council of Europe Framework Convention on Artificial Intelligence (the first binding international AI treaty); the EU AI Act and other regional AI governance frameworks. While technical and industry communities are already engaged through a range of fora - from ISO and ITU standard-setting processes, to multistakeholder mechanisms including the Global Partnership on AI and the OECD AI Policy Observatory, as well as voluntary industry compacts and initiatives - the Global Dialogue can perform an integrating function by translating the outputs of those processes into universally applicable governance commitments, anchored in international human rights. UNICEF-led and co-led initiatives offer concrete vehicles for implementation and good practice, including: the UNICEF Policy Guidance on AI for Children; UNICEF's toolbox for assessing child rights impacts in relation to the digital environment (D-CRIA) and corporate disclosure recommendations on child rights impacts in relation to the digital environment; and the Responsible Data for Children initiative co-led by UNICEF with The GovLab.

The Global Dialogue's credibility depends on the genuine and meaningful participation of the full range of stakeholders envisioned in its mandate. UNICEF welcomes the multistakeholder character of the Dialogue, and the combination of high-level plenaries and structured thematic breakouts co-chaired by member states together with relevant stakeholders. UNICEF commends efforts by the Joint Secretariat to facilitate inclusive participation of representatives and stakeholders of developing countries by offering travel support, through voluntary contributions. UNICEF also welcomes the role of the Independent International Scientific Panel in encouraging inclusive, substantive and evidence-based dialogue. In order to encourage meaningful exchange, practical cooperation, and actionable policy discussions, UNICEF submits the following recommendations for the format and structure of the Dialogue: • A dedicated, structured space for children and young people to contribute to the Dialogue as prepared and supported contributors • Structured multi-stakeholder dialogues in which Member states, international organisations, civil society, academia, private sector and technical experts are seated together at the table • Pairing of the Panel's presentations with facilitated policy dialogue, enabling member states and other stakeholders to discuss practical implications and responses • Problem-solving and co-design sessions anchored in specific AI governance challenges • A dedicated session for presentation of emerging practices, tools and initiatives • An inclusive, well designed hybrid format • Stakeholders from low- and middle-income countries specifically supported to participate • Technical and industry communities engaged not merely as sources of expertise, but as duty-bearers, with participation structured around accountability to agreed normative frameworks rather than deference to technical authority alone • UNICEF strongly supports continuity mechanisms beyond the inaugural Dialogue. We specifically call for a standing mechanism on children's rights and participation within the Global Dialogue framework, or a designated children's rights focal point within the Joint Secretariat.

UNICEF notes a persistent and structural participation deficit for children and young people in global AI governance processes. This is both a rights gap (under Article 12 of the Convention on the Rights of the Child) and a governance gap: the populations likely to be most affected by long-run AI trajectories are those with the least voice in shaping them. Global AI governance efforts should promote the safe, inclusive and meaningful participation of children and young people. A dedicated, structured space should be created for children and young people to contribute to the Global Dialogue and UNICEF stands ready to support the meaningful participation of children and young people in the process, including from contexts where gender norms or safety concerns may otherwise constrain their voice. UNICEF calls on the Joint Secretariat to: • Create dedicated, structured opportunities for children and young people to contribute to the Global Dialogue - not as a symbolic gesture but as substantive contributors with adequate preparation and support • Explore real-time digital participation channels that enable children and young people in different regions and time zones to engage with the Global Dialogue proceedings • Apply child participation principles and safeguards in all consultations involving children, in line with UNICEF's Guidance on Consultations with Young People and the methodology developed for General Comment 27 of the Committee on the Rights of the Child. UNICEF will host a dedicated preparatory consultation with selected global experts in children's rights, safety and wellbeing on 13 May 2026, with the aim of feeding directly into the Global Dialogue discussions. UNICEF is also supporting OHCHR on a linked consultation with children and young people.

The Global Dialogue has an opportunity to model the inclusive, rights-respecting norms it seeks to establish, starting with how it engages. UNICEF recommends the following innovative formats, grounded in its mandate and operational experience: Structured modalities for participation of children and youth: The Dialogue should incorporate dedicated sessions co-designed with, and led by, children and young people. UNICEF has experience facilitating intergenerational policy dialogues and can support formats in which young people present evidence, challenge assumptions, and help shape conclusions. Youth delegates should have floor access equivalent to Member state representatives. Hybrid, deliberative and participatory methods: Hybrid and asynchronous formats could help bring in perspectives that formal negotiating rooms systematically exclude, particularly from low- and middle-income countries, low connectivity environments, and communities most affected by AI harms. The Dialogue should explore participation formats that lower the barriers to meaningful engagement for these groups. While there may also be an opportunity to pilot the use of AI-assisted deliberative tools within the process, any such tools would need to be used responsibly and inclusively, and should be governed by clear safeguards, as well as including human oversight and transparency about how inputs are processed and represented, as well as mechanisms for participants to contest or correct outputs. Cross-constituency roundtables with structured outputs: Small-group dialogues mixing Member states, international organisations, civil society, private sector, academia, technical experts, and rights-holders may produce more substantive results and interactive dialogue than plenary formats. UNICEF recommends designing these around specific implementation challenges and solutions, rather than broad themes.

UNICEF has developed and co-developed a number of practical tools, reports and frameworks that can offer concrete models and guidance for effective, child rights-based AI governance. UNICEF recommends the Global Dialogue identify and amplify such initiatives as part of a shared repository of references, emerging policies and practices in child rights-based AI governance. For example: • UNICEF Policy Guidance on AI for Children, Version 3.0 (2025) provides recommendations for AI policies and systems that uphold child rights. Drawing on the Convention on the Rights of the Child, it sets out ten requirements for child-centred AI and offers a replicable model for embedding children's rights into AI development and regulatory frameworks in both government and private sector (https://www.unicef.org/innocenti/reports/policy-guidance-ai-children) • UNICEF rapid review of AI regulatory frameworks and their implication for child rights (https://www.unicef.org/documents/artificial-intelligence-governance-motion) • UNICEF toolbox for assessing child rights impacts in relation to the digital environment (https://www.unicef.org/childrightsandbusiness/workstreams/responsible-technology/D-CRIA) • UNICEF one-pager on Child Rights and Responsible Technology: Key UNICEF Resources for Businesses (April 2026) - this presents a high-level overview of key resources available with links and further guidance (https://www.unicef.org/childrightsandbusiness/reports/key-resources-responsible-technology) • General Comment 25 of the Committee on the Rights of the Child (2021) provides authoritative guidance on states' obligations in the digital environment, including in relation to AI, and has been used as a reference framework in several national legislative processes. • The Joint UN System Statement on AI and the Rights of the Child (2025) demonstrates a model for multi-agency coordination on children and AI within the UN system, calling on AI governance frameworks to explicitly address children's rights (https://knowledge.unicef.org/knowledge-management/resource/joint-statement-artificial-intelligence-and-rights-child) • UNICEF Generation Unlimited initiative and its digital skills programming across more than 100 countries offers a proven model for closing AI capacity gaps among adolescents, especially in low- and middle-income countries (https://www.unicef.org/kazakhstan/media/3391/file/Gen-U%20-%20Generation%20Unlimited.pdf) • UNICEF guidance on designing and facilitating meaningful participation and consultation processes with children and young people: Methodology for the Participation of Children in CRC General Comment 27 (Child Rights Connect, 2025) UNICEF Guidance on Consultations with Young People (2022); UNICEF spotlight guidance for businesses on engaging with children in impact assessments (2025) • Responsible Data for Children - an initiative by UNICEF and The GovLab to provide guidance, tools, and leadership that supports the responsible handling of data for and about children, including for AI (https://rd4c.org/) • UNICEF guidance on AI in gender-based violence in emergency programming (2025), providing a survivor-centred, feminist framework for assessing AI risks and benefits in contexts where girls face acute protection risks, with applicability beyond emergencies (https://knowledge.unicef.org/child-protection/resource/artificial-intelligence-gender-based-violence-emergency-programming-perils-and-potentials)