Joint Multistakeholder Statement on the UN Global Dialogue on AI Governance

Preamble

We, the members of the Freedom Online Coalition (FOC), welcome the UN Global Dialogue on Artificial Intelligence Governance and its inaugural convening in Geneva in July 2026. Established by General Assembly resolution A/RES/79/325, the Dialogue offers a multistakeholder platform for all relevant stakeholders to exchange on international AI cooperation, share best practices and lessons learned, and promote open, transparent, and inclusive discussions on AI governance.

The FOC is a cross-regional coalition of governments committed to the promotion and protection of human rights and fundamental freedoms, both online and offline, with a strong multistakeholder foundation represented by the FOC Advisory Network (FOC-AN). Noting our Joint Statements on AI and Human Rights (2020, 2025) and on Responsible Government Practices for AI Technologies (2024), we reaffirm that AI governance must be firmly grounded in international law, including international human rights law, and developed through genuine, inclusive multistakeholder processes that centre the voices of those most affected.

This statement sets out the FOC’s priorities across three areas: (I) substantive issues aligned with the Dialogue’s thematic clusters; (II) process and modalities; and (III) the FOC’s contribution to the Dialogue going forward. The statement is intended to inform the Dialogue, and provide a basis for continued work afterwards.

I. Substantive Priorities

A. Opportunities, Implications and Human Rights

The FOC acknowledges that AI systems offer opportunities for human development and innovation. When the entire lifecycle of AI is responsible and complies with international law, including international human rights law, it could generate significant social and economic benefits: improving public health, expanding access to education and public services, strengthening democratic participation, and supporting progress toward the UN Sustainable Development Goals (SDGs). States and other stakeholders play a critical role in enabling and promoting these benefits for all, including for those who have been excluded from the benefits of technological advancements.

At the same time, AI systems can be designed, developed, deployed, or used in ways that pose serious risks to human rights, democracy, and the rule of law. Since the FOC’s 2020 Joint Statement, the scale and urgency of these risks have grown. AI is being used systematically to suppress dissent, enable unlawful surveillance, enable and amplify gender-based violence, and online child sexual exploitation and abuse, spread misinformation and disinformation, and deepen structural inequalities. These harms have cross-border reach with extraterritorial consequences. The governance decisions made now will shape how AI is developed for decades to come. Human rights should be an overarching, cross-cutting foundation embedded across all thematic discussions of the Dialogue.

The Dialogue should underline the need to:

• Refrain from using or facilitating the use of AI systems for repressive purposes, including unlawful or arbitrary surveillance, as well as censorship, or the targeting of civil society, journalists, and human rights defenders.
• Adopt proportionate, human rights-respecting safeguards—including risk assessments and meaningful human oversight, with clear procedures for transparency, redress and evaluation.
• Protect individuals and communities most at risk of AI-enabled harm, particularly children, women and girls, persons belonging to minority communities, Indigenous Peoples, persons with disabilities, LGBTI persons, and others that may be more vulnerable to experiencing harm online, including those facing systemic discrimination. Protections should include representative training data and ongoing bias mitigation.
• Address AI-enabled technology-facilitated gender-based violence, including non-consensual intimate imagery. Further addressing gender-based harms by developing legal, regulatory and accountability measures. 
• Address AI-generated disinformation targeting individuals, through legal, regulatory, technical, and organisational safeguards.
• Address and counter AI-enabled harms against children, including the possession, creation, presentation, dissemination, provision of, or sharing access to online child sexual exploitation and abuse material (CSAM). 
• Support rights to work, an adequate standard of living and social security against the attempted use of AI to replace workers or undermine labor rights.
• Promote cultural and linguistic diversity in AI development, including through multilingual and locally relevant AI models, to promote fair access to the benefits of AI.
• Assess, address and mitigate the environmental footprint of AI, including its energy and resource intensity and its effect on climate change, in a manner consistent with international human rights obligations and the 2030 Agenda.
• Resist approaches to AI governance that tolerate uses of AI incompatible with international human rights law.

B. Bridging Divides: Capacity, Access and Inclusion

The benefits and risks of AI remain profoundly unequally distributed. The Dialogue should take into account and address the inequity in governance and rule-shaping as well as gaps that limit meaningful participation by developing countries, both in AI governance discussions and in access to the benefits of AI. Digital exclusion and inequality are core AI governance and capability concerns, not secondary ones and are not gender neutral. Women and girls often face compounded barriers to meaningful connectivity, affordable devices and digital literacy and skills, further limiting their participation in digital economies and AI governance processes.

Particular attention should also be given to regional and subnational disparities in digital infrastructure, institutional capacities, connectivity digital skills, and access to innovation ecosystems. Effective AI governance requires differentiated capacity-building efforts and territorially informed approaches that reflect the diverse realities and specific development conditions within countries.

The Dialogue should underline the need to:

• Prioritise capacity-building for developing countries, by leveraging existing UN and other multistakeholder mechanisms and initiatives, and support the implementation of national and regional AI strategies, including the African Union’s AI strategy and other regional frameworks, in consultative and rights-respecting ways.
• Recognise data governance as a foundational pillar of AI governance.
• Promote transparent, accountable, and rights-respecting data governance frameworks, including appropriate transparency regarding training data, mechanisms to support the traceability and provenance of AI-generated content, and measures that enable users to identify AI-generated or AI-manipulated content.
• Strengthen coordination with relevant international, multistakeholder processes such as the CSTD Multistakeholder Working Group on Data Governance (WGDG), in line with the Global Digital Compact and the outcomes of the WSIS+20 review, as well as regional efforts such as OECD and APEC.
• Engage national and regional multistakeholder initiatives, including IGF National and Regional Initiatives (NRIs), to channel grassroots and bottom-up perspectives into global policymaking and responsible tech development and deployment.
• Broaden the governance agenda beyond AI benefits and risks to include harms from digital exclusion, inequality, and the underrepresentation of developing countries in international discussions and processes on AI.
• Promote gender-responsive approaches to digital inclusion, including targeted efforts to improve women’s and girls’ access to connectivity, devices and digital literacy, to enable more equitable participation in AI governance and digital economies.
• Promote scientific and technical cooperation mechanisms — including South-South and triangular cooperation — aimed at strengthening national and regional AI capacities, knowledge exchange, research collaboration, and technology-sharing initiatives especially in developing countries.
• Recognize that the benefits, capabilities, and governance of AI remain unevenly distributed, with significant concentrations of market and governance power across parts of the AI lifecycle, including compute infrastructure, data, and foundation models. Promote inclusive, equitable, and participatory approaches that broaden access to AI development and governance, strengthen the meaningful participation of developing countries and underrepresented communities, and support commons-based and community-led models where appropriate.

C. Safe, Secure and Trustworthy AI

The FOC reaffirms its commitment to AI that is safe, secure and trustworthy throughout its full lifecycle. Governments must lead by example, integrating responsible AI practices—including human rights impact assessments, procurement standards, and access to remedies where appropriate—into their own policies and procedures, as set out in our 2024 Joint Statement. Governments should also ensure that AI systems are deployed in a transparent, accountable and predictable manner when used to support the delivery of public services. The private sector also bears responsibilities to respect human rights throughout the AI lifecycle.

Efforts to advance safe, secure, and trustworthy AI should also promote equitable access to technological capabilities, infrastructure, knowledge, and innovation opportunities, particularly for developing countries, so as to avoid deepening existing structural and technological divides.

The Dialogue should underline the need to:

• Advance interoperability and coherence in AI governance approaches, drawing on the OECD Recommendation of the Council on AI, the UNESCO Recommendation on the Ethics of AI, the Council of Europe Framework Convention on AI and Human Rights, Democracy and the Rule of Law, and other relevant instruments, in a human-centric and innovation-friendly manner, respecting international human rights law.
• Promote rights-respecting technical standards, with meaningful input from human rights experts and civil society, and refer in this regard to the FOC’s Joint Statement on Technical Standards and Human Rights in the Context of Digital Technologies.
• Strongly urge AI developers and deployers in the private sector to carry out human rights due diligence across the full AI lifecycle, including by implementing Safety by Design and Secure by Design approaches in platform design and oversight, in line with the UN Guiding Principles on Business and Human Rights, with special attention to high-risk contexts and groups at heightened risk, and encourage transparency through tools such as the Hiroshima AI Process Reporting Framework.
• Recognize the importance of effective technical and legal safeguards — including provenance information, authenticated communications, reporting infrastructure and accountability measures — against large-scale AI-generated disinformation and fraud, prior to deployment of any new AI technology.
• Establish, maintain and enforce privacy and data protection rights for individuals and communities in how their data is collected, stored, used and distributed by AI.
• Leverage existing reporting mechanisms on cross-border AI incidents, including unlawful use or misuse in conflict and disinformation contexts, to enable early warning, shared analysis, and timely response across jurisdictions.
• Ensure that the AI lifecycle respects fair labor practices and the human rights of workers involved in data labeling, model training, and content moderation by requiring supply chain transparency and protecting workers from precarious or hazardous conditions.

D. Transparency, Accountability and Human Oversight

Transparency and accountability are preconditions for public trust in AI systems. States must protect against human rights abuse within their territory and/or jurisdiction by third parties, including business enterprises, which requires taking appropriate steps to prevent, investigate, punish and redress such abuse through effective policies, legislation, regulations and adjudication. Those affected by AI-driven decisions should be able to understand, challenge, and seek remedy for those decisions.

The Dialogue should underline the need to:

• Promote transparency and ensure meaningful, effective human oversight, where appropriate to the context, of AI systems, particularly in high-risk areas such as law enforcement, the justice system, border management, and public benefits administration, consistent with states’ international obligations. Oversight mechanisms should be capable of identifying, preventing and addressing harms, including those that disproportionately affect vulnerable groups.
• Ensure effective access to remedy for individuals harmed by AI systems, as appropriate, such as through timely human review, notification of adverse automated decisions, and well-functioning, accessible and enforceable redress protocols and mechanisms. This should support safe, inclusive and rights-respecting outcomes and enable individuals to challenge and seek resolution for harmful or unfair impacts.
• Encourage providing effective access to remedy, including through human-in-the-loop requirements for high-impact decisions and establishment of independent national supervisory authorities.
• Recognise the role of OHCHR in advising on global AI governance and human rights. Strengthening its role in the Global Dialogue and the Independent International Scientific Panel on AI, and supporting capacity initiatives such as the Digital Rights Advisory Service.
• Encourage the use of AI to advance gender equality, the SDGs, and human rights, including through knowledge exchange at forums such as the IGF and the ITU AI for Good Global Summit.
• Promote and enable independent researchers’ ability to study and test AI systems’ adherence to human rights principles and the UN Guiding Principles on Business and Human Rights.
• Encourage the responsible use of AI to support biodiversity protection, climate change adaptation, environmental monitoring, disaster-risk management, and evidence-based public policies aimed at advancing sustainable development.

II. Process: Multistakeholder, Inclusive, Coherent

The FOC has long championed open, inclusive, and multistakeholder approaches to Internet and digital governance, and considers these principles indispensable to the Dialogue’s legitimacy and effectiveness. The FOC highlights the role of the Dialogue as a connector and coordinator across the broader UN AI governance landscape, including the Global Digital Compact, WSIS, IGF, CSTD, UNESCO and regional bodies, rather than creating parallel structures or duplicating existing work, in line with broader UN reform efforts.

III. The FOC’s Contribution to the Dialogue

The FOC brings to this Dialogue more than five years of sustained normative work on AI and human rights, a demonstrated commitment to multistakeholder governance, and the practical experience of its member States in implementing responsible AI policies at the national level. We consider our comparative advantage to rest on our grounding in international law, including international human rights law, our inclusive governance model — encompassing both government members and a non-governmental advisory network — and our capacity to translate principles into concrete policy commitments.

The FOC commits to actively engaging in all phases of the Dialogue, sharing the practical experience of member States, continuing to develop our normative guidance on AI and human rights, and advocating for the meaningful participation of civil society and affected communities throughout the process. We will also pursue coalition expansion and cross-regional partnerships to broaden the reach and impact of rights-respecting AI governance approaches globally.

We call on all States and stakeholders — from the private sector, civil society, academia, youth and the technical community — to engage constructively in the Dialogue, to uphold the multistakeholder spirit in which it was conceived, and to ensure that AI governance decisions realize the benefits of AI for all and protect the human rights and fundamental freedoms of all.*

*The Republic of Korea dissociates itself from the statement.

The Republic of Korea informed the Chair during the silence procedure that it wished to dissociate itself from the statement due to national considerations, while confirming that it did not seek to break silence. Having considered the request, the Chair agreed, in the particular circumstances of this case, to record the Republic of Korea’s position in a footnote without reopening the silence procedure. This should not be understood as modifying or supplementing the procedural framework set out in the FOC Terms of Reference.

Endorsed by the following non-governmental stakeholders:

FOC Advisory Network (2026-2028 cohort):

• Alberto Fernandez Gibaja, International IDEA
• Alexandria Walden, Google
• Edetaen Ojo, Media Rights Agenda
• Elonnai Hickok, Global Network Initiative
• Flavia Alves, Meta
• J. Carlos Lara, Derechos Digitales
• Jennifer Chung, DotAsia Organisation
• Jordi Calvet-Bademunt, Vanderbilt University
• Khadijah El-Usman, Paradigm Initiative
• Konstantinos Komaitis, Individual capacity
• Kristina Rikhter, Association Human Constanta International
• Layal Jebran, Individual capacity
• Liz Orembo, Research ICT Africa
• Mahsa Alimardani, Individual capacity
• Marlena Wisniak, ECNL
• Michael Samway, Individual capacity
• Mishi Choudhary, Software Freedom Law Center (sflc.in)
• Mohamed Farahat, Individual capacity
• Nicole Karlebach, Microsoft
• Owen Bennett, Individual capacity
• Pablo Hinojosa, Individual capacity
• Patrick Day, Cloudflare
• Patryk Pawlak, European University
• Rafael A. F. Zanatta, Data Privacy Brasil
• Sabhanaz Rashid Diya, Tech Global Institute
• Sara Fratti, Iniciativa Latinoamericana por los Datos Abiertos
• Sarah Clarke, IFEX
• Sophia Tawonga Longwe, Individual capacity
• Timea Suto, ICC-Basis
• Zach Lampell, International Center for Not-for-Profit Law

Former FOC Advisory Network Members from the 2024-2026 cohort:

• Amalia Toledo, Lead Public Policy Specialist
• Association for Progressive Communications (APC)
• Center for Democracy & Technology (CDT)
• Dr. Katharine M Millar, Associate Professor of International Relations at the London School of Economics and Political Science, UK
• Dr. Sebastian Smart, Senior Researcher at Anglia Ruskin University, Cambridge, UK