OHCHR

The inaugural Global Dialogue on AI Governance comes at an opportune time as the world is being transformed by AI, but the necessary governance to ensure that AI respects human rights, promotes sustainable development, and benefits all of humanity remains underdeveloped. Firstly, a successful outcome for the Dialogue could include a global commitment towards governing AI that is fully grounded in the UN Charter, ensuring that the normative direction for AI Governance respects, protects and promotes human rights and accelerates sustainable development. Secondly, to turn commitments into action, multistakeholder collaboration platforms that can take forward some commitments and turn principles into more granular actions and recommendations would help close some of the governance, capacity, and knowledge gaps on AI.

• AI capacity-building
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Protection and promotion of human rights
• Transparency, accountability, and human oversight

All thematic areas of the Global Dialogue on AI are important from a human rights perspective. For AI governance, the protection and promotion of human rights as well as transparency, accountability, and human oversight serve as the foundational building blocks for any AI governance framework that is grounded in international law. Understanding the social, economic, ethical, cultural and linguistic, and technical implications of AI will enable a better understanding of how AI can realise or adversely affect a wide range of human rights. AI capacity building is essential to enable Member States to govern AI effectively, reduce cross-country inequalities, and realize their right to development.

Digital divides and differences in capacity represent key challenges that undermine the equal enjoyment of human rights. How to overcome these inequities should be a core topic of the Dialogue. The GDC mentions several aspects of AI, such as when applied to digital public goods and digital public infrastructure, open AI models, AI technical standards, and data privacy. GDC also includes a chapter to enhance international governance of AI for the benefit of humanity. OHCHR understands that the governance of AI is closely linked to human rights. Across all the new mechanisms or processes set up under the GDC, human rights should be central, especially in the context of AI governance. As the High Commissioner stated: "Human rights provide an inspiring vision of AI that can make our world safer and more peaceful, and develop it through human creativity, imagination and purpose. In short, human rights provide a governance model that is long-term and inter-generational, that links our past with our present and our future." Human rights are a cross-cutting theme. The GDC implementation, including the Global Dialogue, should push for the development of AI for digital public goods and foster AI technical standards that embed a human rights approach. Implementation measures should additionally consider key themes like data privacy, cross-border data flows, freedom of expression, access to information, surveillance, digital public infrastructure, bias/discrimination, among other human rights-related aspects of the GDC. Environmental sustainability, including the right to a clean, healthy, and sustainable environment recognized in A/RES/76/300, is an emerging issue absent from the listed themes that merits significant attention. The expansion of AI infrastructure and capacity poses concrete risks to fragile ecosystems and to local communities' access to water.

• Children's rights and safety. CRC and General Comment No. 25 provide a comprehensive framework for live policy debates on balancing protection, access, and empowerment; The 2026 Joint Statement on AI and the Rights of the Child (13 UN entities) addresses core issues. • Labour markets and economic disruption. As job displacement accelerates, grounding the response in human dignity and ICESCR (Arts. 6, 7, 9) is essential to a just transition. • Equitable access and shared benefit from AI. Right to share in scientific advancement (UDHR, 27; ICESCR, 15) reframes AI access in existing law. Concentrated AI infrastructure and expertise also implicate the Right to Development. • AI buildout, resources, and the right to a healthy environment. AI's physical infrastructure generates significant, unevenly distributed impacts on water, energy, land, and community rights, often in already stressed regions. • Agentic AI and meaningful human oversight. Increasingly autonomous systems are widening accountability gaps. The rights to effective remedy, due process, access to justice, rule of law, and meaningful human oversight are immediately implicated. • Non-discrimination and algorithmic classification. AI systems that score, sort, and profile people produce discriminatory outcomes. ICCPR (26), ICERD, CEDAW, CRPD, and the UNGPs/B-Tech due diligence framework provide applicable standards. • Gender. AI is reproducing and amplifying gender-based discrimination. The CEDAW framework and the WG's forthcoming HRC report on AI and gender equality (06/2026) provide governance foundations. • Information integrity & Freedom of expression. AI-generated disinformation and AI-powered content moderation threaten the information ecosystem. The ICCPR Art. 19 three-part test provides the basis for assessing justified responses. • Data and privacy. AI depends on massive data extraction. ICCPR Art. 17 provides a universal basis for data governance with a built-in proportionality test. • Interoperability and AI Standards. Alignment of technical standards with universal human rights values is essential (GDC, 58). The Seoul Statement (ISO, ITU and IEC, 2025), explicitly references human rights.

The Dialogue can serve two complementary functions. First, as a cross-sectoral convening that builds and activates coalitions to advance AI governance - with sustained multistakeholder engagement required to carry that agenda beyond the room. Second, as a bottom-up process that creates space for those most affected by AI to place responsible use and governance on the agenda. Additionally, the Dialogue can also serve as a platform for multistakeholder coalitions to continue to build on the objectives of the Dialogue, including through intersessional work, to ensure that each iteration of the Dialogue further builds on previous work.

The UN human rights system offers a substantive body of norms, standards, and accountability mechanisms that the AI Dialogue should actively draw upon. The Human Rights Council, its Special Procedures - including Special Rapporteurs, independent experts, and Working Groups - and the UN Treaty Bodies, which monitor State compliance with international human rights law, have collectively produced authoritative guidance directly relevant to AI governance. This body of work reflects binding legal obligations that Member States have already accepted, providing the Dialogue with a principled, universally grounded framework. Key examples include: • OHCHR reports on human rights and emerging digital technologies (A/HRC/56/45), privacy and AI (A/HRC/48/31), technical standards and human rights (A/HRC/53/42), AI, privacy and discrimination (A/HRC/60/45) and countering disinformation (A/77/287). • The B-Tech Project, which has produced practical human rights due diligence guidance for tech companies on AI and Generative AI. • Special Procedures reports covering freedom of expression and AI, hate speech, disinformation, social media governance, and the rights of persons with disabilities. • Treaty Body outputs, including General Comments on children's rights in the digital environment (CRC/C/GC/25) and racial profiling (CERD/C/GC/36). • General Assembly and HRC resolutions calling for bans on AI applications incompatible with international human rights law, human rights due diligence (A/RES/78/213), and the right to privacy (A/RES/77/211). Connecting the Dialogue to these mechanisms would also strengthen accountability over time: human rights bodies can flag emerging AI-related concerns, monitor implementation of governance commitments, and ensure affected communities have a voice in norm development. Further examples are provided in OHCHR's response to Question 18.

• To seize the opportunities of the Global Dialogue on AI Governance, it is critical that stakeholders engage with this UN process and that a genuine exchange of views can occur. OHCHR recommend that the organizers of the Global Dialogue on AI a) Allocate sufficient time and speaking opportunities to stakeholders, fostering a real dialogue between Member States and non-governmental actors, ensuring sufficient speaking time with minimum 50% for non-governmental stakeholders
• b) Continue, improve, and foster the facilitation process with Participation Support for the AI Dialogue, focusing on participants from Global South, including civil society organizations
• c) Ensure they take into consideration civil society submissions/proposals in the Co-Chair's summary document
• d) implement a Multistakeholder Advisory Group for the Global Dialogue on AI Governance, with members from governments, the private sector, media, civil society, and the technical community, and building upon the experience from the MAG-IGF and the WSIS Informal Multistakeholder Sounding Board .

Several groups remain consistently underrepresented in global AI governance discussions: civil society and affected communities from the Global South, indigenous peoples, persons with disabilities, women and girls, older persons, and those experiencing poverty or marginalization. These are often the populations most directly affected by AI systems, yet least present in the forums shaping governance responses. Structural and financial barriers are the primary obstacle. Meaningful inclusion requires targeted support - covering travel, interpretation, and capacity - alongside transparent eligibility and selection criteria. The Participation Support initiative for the Geneva 2026 Dialogue is a welcome step in this direction and should be extended to future Dialogue sessions in both Geneva and New York, with public disclosure of donor contributions and participant composition. Beyond funding, process design matters. The NetMundial+10 Sao Paulo Guideline, recognized in WSIS+20 outcome resolution, provide tested procedural guidance for open and inclusive multistakeholder processes and should inform how the Dialogue is structured moving forward. OHCHR also recommends that the Dialogue draw on established good practices from other UN digital governance processes, which have developed concrete modalities for civil society participation. The April 2026 civil society letter to the Dialogue Co-Chairs on procedural recommendations offers additional actionable guidance that merits serious consideration .

To promote dialogue, interactive thematic discussions should be emphasised, enabling speakers to engage with ideas presented and with each other's statements. Linear meetings where speakers make sequential pre-prepared statements would detract from the ability of participants to engage in a dialogue. Meaningful dialogue requires moving away from sequential, pre-prepared statements toward formats that generate genuine exchange. OHCHR recommends the following approaches, drawn from good practices in UN and multistakeholder processes: Interactive thematic panels: rather than keynote style presentations, we recommend that structured panels encourage speakers to respond directly to each other's positions and questions from the audience. Moderators can have a mandate to push for specificity. Breakout and working group sessions: we recommend breakout and working group sessions, keeping multistakeholder participation and fostering the dialogue between civil society, companies, technical community and governments. A rapporteur for each of these group sessions could feed and present the main points discussed in the Plenary session. Problem-based roundtables: Structured sessions to brainstorm solutions, identify key issues and explore possible courses of action in response to a given problem can be useful to bring different perspectives to the same issue. We recommend this format to foster an open dialogue and collaborative problem-solving. Asynchronous and remote participation: real-time online and asynchronous participation should be fostered, allowing Global South civil society organizations to participate meaningfully even if travel costs are a barrier to in-person participation. Transparent documentation: We recommend that all different tracks from the AI Dialogue maintain transparent processes, with full documentation and public available on the AI Dialogue website (Examples: WSIS+20 and Ad Hoc Committee on Cybercrime).

• Mapping report regarding human rights and new and emerging digital technologies (NET) (A/HRC/56/45). • 2021 Report on the right to privacy in the digital age A/HRC/48/31, focusing on human rights and AI. • 2022 Report on the right to privacy in the digital age A/HRC/51/17, addressing spyware, encryption, public surveillance. • 2022 Report on internet shutdowns, A/HRC/50/55; • 2023 Report on technical standards and human rights, A/HRC/53/42. • 2023 Study on "Digital Border Governance: a human rights-based approach". • 2025 Study on "New pathways for incorporating international human rights into standards development for digital technologies". • 2020 SG report on the role of new technologies for the realization of economic, social and cultural rights, drafted by OHCHR, A/HRC/43/29. • 2022 SG report on countering disinformation, drafted by OHCHR, A/77/287. • OHCHR's B-Tech project has produced practical guidance on human rights due diligence for tech companies in their development and use of AI, Generative AI, among many other topics (Generative AI project). • Reports of the HRC Special Procedures, covering a vast range of topics, e.g. freedom of expression and AI (including hate speech, disinformation and social media governance); freedom of thought; rights of older persons and persons with disabilities (including assistive technologies). • The Human Rights Treaty Bodies have produced General Comments and General Recommendations focusing on digital technology and specific rights including children's rights in the digital realm, the rights to non-discrimination, etc, e.g., Committee on the CERD concerning Racial Profiling; Committee on the CRC on children's rights in relation to the digital environment. • General Assembly resolutions, e.g. calling for bans on AI applications that are impossible to operate in compliance with IHRL, and for human rights due diligence (78/213); the right to privacy, (77/211); and the HRC, e.g., right to privacy in the digital age, 54/21; NET, 53/29; promotion and protection of human rights on the internet (57/29).