UNAIVERSE

The International Telecommunication Union convened the first AI for Good Global Summit in Geneva with UN partners, government officials, researchers, civil society and technology companies. The summit was designed to identify practical uses of AI that could accelerate the Sustainable Development Goals and to connect technical experts with organizations working on global problems. It subsequently developed into the year-round AI for Good platform and a recurring global summit led by ITU in partnership with numerous UN entities.

The United Nations Interregional Crime and Justice Research Institute established its Centre for Artificial Intelligence and Robotics in The Hague. The Centre examines AI and robotics from the perspective of crime prevention, criminal justice, security, human rights and the rule of law. Its work has included malicious uses of AI, responsible innovation, law-enforcement applications and international cooperation.

Secretary-General António Guterres issued the Roadmap for Digital Cooperation as a follow-up to the High-level Panel on Digital Cooperation. The Roadmap addressed universal connectivity, digital inclusion, digital public goods, digital capacity-building, human rights, digital trust and security, and global cooperation on artificial intelligence. It proposed strengthening international AI cooperation and considering a multi-stakeholder advisory body to examine global AI governance.

The World Health Organization published Ethics and governance of artificial intelligence for health. The guidance sets out six principles: protecting human autonomy; promoting human well-being, safety and the public interest; ensuring transparency and explainability; fostering responsibility and accountability; ensuring inclusiveness and equity; and promoting responsive and sustainable AI. It provides recommendations for governments, developers, health providers, researchers and funders.

UNESCO’s General Conference adopted the Recommendation on the Ethics of Artificial Intelligence at its forty-first session. UNESCO describes it as the first global normative instrument on AI ethics and states that it was adopted by its 193 Member States. The Recommendation covers human rights, dignity, fairness, transparency, accountability, privacy, data governance, education, labour, culture, health, the environment and international cooperation. It also calls for implementation tools, national reporting and institutional capacity.

States Parties to the Convention on Certain Conventional Weapons began informal expert discussions on lethal autonomous weapons systems in 2014. A formal Group of Governmental Experts was established in 2016 and began meeting in 2017. The GGE works by consensus and has agreed guiding principles, including that international humanitarian law applies fully and that human responsibility for decisions on weapon use must be retained.

ITU-T develops standards and pre-standardization work for AI and machine learning in telecommunications, networks, multimedia, digital health, autonomous driving and environmental efficiency. Examples include ITU-T Y.3172 on integrating machine learning into future networks, Y.3173 on evaluating intelligence levels and Y.3174 on data handling for machine learning. ITU also supports focus groups and joint initiatives with WHO and other organizations.

UNESCO developed the Readiness Assessment Methodology to help governments evaluate their institutional, legal, social, economic, scientific and technical readiness to implement the 2021 Recommendation. It also developed an Ethical Impact Assessment tool for evaluating AI systems and policies. Country profiles and completed assessments are published through the Global AI Ethics and Governance Observatory.

The Human Rights Council adopted, without a vote, a resolution on new and emerging digital technologies and human rights at its forty-first session. It recognized that technologies can facilitate human rights while also creating or amplifying violations and abuses. The resolution requested work by OHCHR and the Council’s Advisory Committee and became part of a continuing Council track on digital technologies.

The Secretary-General’s Data Strategy aims to strengthen data governance, capabilities, analytics and responsible data use across the UN. It promotes better decision-making, stronger programme delivery and more effective organizational management. AI adoption in the Secretariat depends on the data governance, quality and skills foundations established by this strategy.

OHCHR’s B-Tech Project applies the UN Guiding Principles on Business and Human Rights to digital technologies and technology companies. Its work addresses business models, human-rights due diligence, remedy, responsible investment and corporate accountability. B-Tech has produced material specifically addressing generative AI harms and responsible company practice.

UNICEF published successive editions of policy guidance on AI and children. The guidance calls for children’s safety, privacy, non-discrimination, transparency, accountability, inclusion, best interests and meaningful participation to be embedded in AI policy and product design. The 2025 edition expands coverage of generative AI, AI companions, environmental impacts, abuse material, supply-chain risks and armed conflict.

UN 2.0 groups data, digital, innovation, behavioural science and strategic foresight as the “Quintet of Change.” The initiative is intended to modernize UN capabilities and make programmes more adaptive, evidence-based and future-oriented. AI is treated as part of the digital and data transformation rather than as an isolated reform stream.

The Department of Peace Operations developed a strategy for digital transformation of UN peacekeeping. It emphasizes stronger situational awareness, protection of civilians, information integrity, data-driven planning, early warning and mission support. AI, analytics, remote sensing and automation may support these goals, subject to operational, legal, ethical and security safeguards.

The Open-ended Working Group on security of and in the use of information and communications technologies addresses international law, responsible State behaviour, confidence-building, capacity-building and cyber threats. AI intersects with this agenda through AI-enabled cyber operations, automated vulnerability discovery, synthetic influence campaigns and defensive applications. The OEWG is not an AI-governance forum, but its norms are relevant to AI-enabled cyber risk.

OHCHR’s privacy-in-the-digital-age work increasingly addresses algorithmic decision-making, surveillance and artificial intelligence. OHCHR has emphasized that States must conduct human-rights due diligence when procuring, developing or using AI systems and must provide safeguards, transparency and access to remedy. This work is connected to recurring General Assembly and Human Rights Council resolutions on privacy in the digital age.

Unite Aware integrates operational information to support situational awareness and decision-making in peacekeeping environments. It is designed to combine information from different mission systems and improve a common operational picture. Public documentation does not provide a complete inventory of any machine-learning models that may be used within specific deployments.

The UN System Chief Executives Board endorsed ten principles for ethical AI use across the UN system. They address proportionality and do no harm, safety and security, fairness, sustainability, privacy, human autonomy and oversight, transparency, explainability, responsibility, accountability, inclusion and participation. The principles apply across the AI lifecycle and to UN entities using or developing AI.

UNDP and ODET support safe, inclusive and interoperable digital public infrastructure for identity, payments and data exchange. The Global Digital Compact encourages digital public goods and open-source technologies. ODET has also examined conditions under which AI systems may qualify as digital public goods.

The Human Rights Council requested an expert study on the human-rights implications of new and emerging technologies in the military domain. The work includes AI and autonomous systems and considers the relationship between technological development, international human-rights law and international humanitarian law.

UN entities use predictive analytics, satellite imagery and machine-learning techniques for food-security monitoring, disaster mapping, displacement analysis and anticipatory action. Public examples include WFP’s SKAI satellite-image project, UNHCR’s Project Jetson experiment and work by OCHA’s Centre for Humanitarian Data. Governance is entity-by-entity: no single UN-wide policy covering all predictive or generative AI use in humanitarian operations was found in the public record as of 28 July 2026.

Guidance on staff use of generative AI was first issued to Secretariat personnel in June 2023 and revised on 5 March 2025 as Updated UN Guidance on the Use of AI Tools at Work. It designates an authorized enterprise chat solution for day-to-day work, directs staff away from unapproved tools for official tasks, and addresses confidentiality, personal data, verification, intellectual property, bias and human accountability. Alongside it, the CEB endorsed a system-wide framework for a model policy on the responsible use of AI and published a report on operational AI use across the system. This guidance sits below the level of a Secretary-General’s bulletin or administrative instruction; no such formal issuance dedicated to generative AI was located in the public policy repositories reviewed.

UN entities have piloted chatbots, virtual assistants, automated translation, satellite-image analysis, anomaly detection and decision-support tools. Public examples are available through UNHCR, WFP, WHO and UNDP. Documentation is uneven, and many pilots are described through innovation pages rather than formal impact evaluations.

The UN frames AI as a technology capable of supporting health, education, agriculture, climate action, public administration and crisis response while also generating risks for equality, rights, employment and the environment. The two 2024 General Assembly resolutions embed this dual framing in intergovernmental policy. ITU’s AI for Good platform and UNDP’s capacity-building work are major implementation channels.

OICT supports enterprise platforms, cloud services, cybersecurity, analytics and digital innovation across the Secretariat. An OIOS evaluation of digital transformation reported a large, decentralized portfolio of digital initiatives across Secretariat entities, including some involving advanced analytics and AI. A complete public register of AI systems, models, vendors, risk classifications and impact assessments was not located.

The United Nations Institute for Disarmament Research conducts research and convenes discussions on AI, military applications, strategic stability, human control, testing, assurance, ethics and responsible innovation. Its Security and Technology Programme examines AI together with cyber capabilities, autonomy and other emerging technologies. UNIDIR’s research supports diplomats but does not itself create binding rules.

WHO, ITU and WIPO established the Global Initiative on AI for Health as a platform for standards, evaluation, governance and implementation. It builds on the earlier ITU-WHO Focus Group on AI for Health. The initiative aims to support countries and stakeholders in evaluating and deploying health AI safely and equitably.

ITU convened the 2023 summit in Geneva, focusing on generative AI, robotics, health, climate, accessibility and the Sustainable Development Goals. The programme combined demonstrations with policy and standards discussions. It marked the platform’s return to a major in-person summit after pandemic-era virtual activity.

The United Kingdom convened the first formal Security Council meeting devoted specifically to artificial intelligence under the agenda item “Maintenance of international peace and security.” Secretary-General António Guterres warned that AI could support development, humanitarian action and peace efforts but could also amplify disinformation, cyberattacks and weapons risks. He supported consideration of a global AI governance institution able to provide scientific assessment and policy coordination. Briefers included Jack Clark of Anthropic and Yi Zeng of the Chinese Academy of Sciences.

The International Labour Organization assessed occupational exposure to generative AI across countries and income groups. It found that clerical work had the highest exposure and concluded that job transformation was more likely than full job elimination. The report also identified substantial gender differences in exposure because women are overrepresented in clerical work in many economies.

The Secretary-General and the President of the International Committee of the Red Cross jointly called on States to negotiate a legally binding instrument establishing prohibitions and restrictions on autonomous weapons systems and to conclude negotiations by 2026. They called for prohibiting unpredictable systems and systems designed or used to apply force against persons, while regulating other systems through limits and meaningful human control.

The Secretary-General launched a 39-member High-level Advisory Body on Artificial Intelligence for a one-year mandate. Its members served in their personal capacities and came from government, industry, academia and civil society across 33 countries. The Body was asked to advance recommendations on global AI governance, improve shared understanding of AI risks and opportunities, and support use of AI for the Sustainable Development Goals.

The Body comprised 39 experts from 33 countries and was described by the UN as geographically diverse and gender-balanced. Its co-chairs were Carme Artigas of Spain and James Manyika of Google-Alphabet. The complete official roster and member biographies are maintained on the Advisory Body website.

The High-level Advisory Body released its interim report, Governing AI for Humanity. The report proposed guiding principles and identified functions that international AI governance should perform, including scientific assessment, risk monitoring, policy interoperability, capacity-building and accountability. It was followed by public consultations and further expert work before the final report.

The General Assembly adopted its first resolution devoted to lethal autonomous weapons systems by a recorded vote of 152 in favour, 4 against and 11 abstentions, at its fiftieth plenary meeting. The resolution requested the Secretary-General to seek views from States and relevant stakeholders and submit a report to the Assembly. It recognized the continuing work of the Convention on Certain Conventional Weapons Group of Governmental Experts while opening a universal General Assembly track.

UN reports describe AI inequality in terms of connectivity, compute, data, financing, skills, language coverage and institutional capacity. The Advisory Body, Global Digital Compact, UNCTAD and ODET call for international cooperation and public-interest infrastructure to prevent concentration of AI capabilities and benefits. Capacity-building for developing countries is therefore treated as a core governance issue rather than a separate development programme.

ODET supports follow-up and implementation through mapping commitments, convening stakeholders, tracking initiatives and supporting intergovernmental processes. The AI commitments were implemented first through negotiations that led to A/RES/79/325, followed by creation of the Scientific Panel and Global Dialogue. The Compact mandates a high-level review during the General Assembly’s eighty-second session.

OCHA’s Centre for Humanitarian Data supports responsible data management, predictive analytics and model-related risk management in humanitarian settings. OCHA also supports anticipatory action based on forecasts and predefined triggers. The data-responsibility framework is public; a single consolidated public OCHA policy governing all generative-AI use was not found in the public record as of 28 July 2026.

The Global Digital Compact calls for interoperable and rights-respecting approaches to data governance. It connects data to privacy, development, cross-border flows, public-interest uses and AI systems. As of 28 July 2026 the Compact had not produced a single binding global data-governance framework; the work remains a set of parallel processes rather than one instrument.

Relevant structures include the Scientific Advisory Board, ODET, the UN 2.0 network, the former High-level Advisory Body on AI and the Independent International Scientific Panel. Their roles are different: internal strategic advice, Secretariat coordination, organizational transformation, time-limited policy recommendations and independent scientific assessment. They should not be presented as a single hierarchy.

UNDP worked with the Italian G7 presidency on an AI Hub for Sustainable Development, initially focused on African ecosystems. The initiative maps needs and opportunities and aims to support partnerships, local innovation, skills, responsible governance and access to resources. It reflects the capacity-building priorities later emphasized in General Assembly resolutions.

UNDP’s Artificial Intelligence Landscape Assessment helps governments assess the national environment for responsible AI. It examines data, infrastructure, institutions, skills, innovation, inclusion and governance. UNDP also offers joint country support with UNESCO, combining UNDP’s landscape assessment with UNESCO’s ethics-focused Readiness Assessment Methodology.

The World Food Programme supports AI applications for satellite-image analysis, food-security monitoring, assistance integrity, logistics and mission planning. Public projects include SKAI, anomaly detection for assistance delivery, an enterprise deduplication service and an AI-supported mission planner. WFP presents AI as a way to increase speed, targeting and operational efficiency while requiring responsible use.

WHO released guidance on the ethics and governance of large multi-modal models used in health care, public health, scientific research and drug development. It recommends regulatory oversight, transparency, independent auditing, cybersecurity, stakeholder participation and post-deployment monitoring. WHO warns that plausible but inaccurate outputs can create serious harm in health settings.

High Commissioner for Human Rights Volker Türk stated that human rights must guide the entire lifecycle of generative AI systems. He emphasized transparency, accountability, non-discrimination, participation, privacy and remedy. He also linked government obligations and corporate responsibilities to the UN Guiding Principles on Business and Human Rights.

The General Assembly adopted by consensus and without a vote the United States-led resolution Seizing the opportunities of safe, secure and trustworthy artificial intelligence systems for sustainable development. The draft was issued as A/78/L.49 and adopted at the Assembly’s sixty-third plenary meeting. The resolution emphasizes international law, human rights, privacy, safety, inclusion and the need to close digital and development divides.

The 2024 summit brought together governments, companies, researchers, civil society and UN organizations. Major themes included AI governance, technical standards, robotics, health, climate and accessibility. It took place during the same year as the first two General Assembly AI resolutions and the final Advisory Body report.

The General Assembly adopted without a vote the China-led resolution Enhancing international cooperation on capacity-building of artificial intelligence. The draft was issued as A/78/L.86 and adopted at the ninety-seventh plenary meeting. It calls for cooperation on infrastructure, financing, knowledge-sharing, technical assistance, data, skills and institutional capacity, with particular attention to developing countries.

The report compiled submissions from Member States, observer States, international and regional organizations, the ICRC, civil society, academia and industry pursuant to A/RES/78/241. It presents differing views on legal sufficiency, human control, prohibitions, restrictions and the appropriate negotiating forum. The report does not itself establish a negotiated UN position.

The Advisory Body released its final report after extensive consultations. It proposed seven measures: an international scientific panel on AI; an annual policy dialogue; an AI standards exchange; a capacity-development network; a global fund for AI; a global AI data framework; and a small AI office within the UN Secretariat. The report also called for global AI governance grounded in the UN Charter, international law, human rights and sustainable development.

Heads of State and Government adopted the Pact for the Future at the Summit of the Future. The Global Digital Compact appears as Annex I and covers connectivity, digital inclusion, human rights online, data governance, digital public infrastructure, trust and safety, and AI governance. The Compact calls for establishment of an Independent International Scientific Panel on AI and a Global Dialogue on AI Governance. It also mandates a high-level review during the General Assembly’s eighty-second session.

The Compact addresses universal connectivity, digital inclusion, digital literacy, human rights online, Internet governance, digital trust and safety, data governance, digital public goods and digital public infrastructure. It reaffirms multi-stakeholder cooperation and links digital transformation to the Sustainable Development Goals. Its scope is explicitly non-military.

The United States convened a high-level Security Council briefing during its December presidency on artificial intelligence and the maintenance of international peace and security. It was the Council’s 9821st meeting. The discussion covered military applications, autonomous weapons, cyber risks, disinformation, strategic stability, international law and human control. The Secretary-General told the Council that AI was outpacing the world’s ability to govern it and urged Member States to agree international guard-rails.

The General Assembly adopted Artificial intelligence in the military domain and its implications for international peace and security by a recorded vote of 159 in favour, 2 against and 5 abstentions, at its fifty-fifth plenary meeting. The resolution addresses military AI beyond the narrower lethal autonomous weapons track. It requested the Secretary-General to seek views and prepare a report cataloguing opportunities, risks, challenges and existing or emerging normative proposals.

ITU’s AI Standards Exchange provides access to standards and related technical publications from ITU, ISO, IEC and other standards-development organizations. It also includes standards news and capacity-building material. The initiative responds to Global Digital Compact commitments on interoperable standards and international cooperation.

The Office of the Secretary-General’s Envoy on Technology transitioned into the United Nations Office for Digital and Emerging Technologies. ODET supports implementation and follow-up of the Global Digital Compact and coordinates work on AI governance, connectivity, digital inclusion, data governance, digital public infrastructure and emerging technologies. The Office also provides substantive and secretariat support to the UN’s new AI governance mechanisms.

The twenty-year review of the World Summit on the Information Society examines connectivity, inclusion, Internet governance and implementation of the WSIS action lines. AI has become increasingly integrated into WSIS discussions. The 2026 Global Dialogue was held alongside the WSIS Forum in Geneva, illustrating growing alignment between the WSIS and Global Digital Compact processes.

ILO and NASK published a refined global index measuring occupational exposure to generative AI. The updated method classified tasks and occupations in greater detail and again found that transformation was more likely than full automation for most jobs. The study estimated that roughly one in four workers globally was in an occupation with some degree of exposure.

The 2025 summit continued ITU’s programme on standards, skills, innovation, robotics and partnerships. It occurred while Member States were finalizing the terms of reference for the Scientific Panel and Global Dialogue. The summit also supported the expanding AI Standards Exchange.

UNHCR published an organizational approach to AI covering refugee protection, assistance, policy, emergency response and internal efficiency. It emphasizes human oversight, safety, accountability, humanitarian principles, participation, data protection and procedural fairness. It also addresses information integrity and the risks of bias or exclusion affecting refugees, asylum-seekers and stateless people.

The General Assembly adopted without a vote the resolution establishing the Independent International Scientific Panel on Artificial Intelligence and the Global Dialogue on Artificial Intelligence Governance. The draft resolution was A/79/L.118, and the adoption meeting record was A/79/PV.89. The resolution provides detailed terms of reference and modalities, including a 40-member independent panel, geographical balance, annual scientific reporting and an inclusive annual dialogue. The Panel’s mandate is limited to the non-military domain.

The Republic of Korea convened an open debate on AI and international peace and security during its Security Council presidency. It was the Council’s 10005th meeting, and it continued over two resumptions. The meeting examined conflict prevention, military uses, human oversight, disinformation, cyber threats, international law and inequality in access to AI capabilities.

The Secretary-General addressed a high-level multi-stakeholder informal meeting launching preparations for the Global Dialogue on AI Governance. He described the Dialogue as a venue for building safe, secure and trustworthy AI grounded in international law and human rights, promoting interoperability and expanding participation in AI governance. The meeting initiated political and stakeholder preparations for the first substantive session.

The General Assembly appointed the Panel’s 40 members on 12 February 2026, by a recorded vote, for a three-year term running to 11 February 2029. The members were drawn from more than 2,600 applications from over 140 countries, following independent review by ITU, ODET and UNESCO, and comprise 19 women and 21 men from all five UN regions. At the Panel’s first plenary meeting on 3 March 2026 the members elected Yoshua Bengio of Canada and Maria Ressa of the Philippines as co-chairs. Members serve in their personal capacity rather than as representatives of governments, employers or institutions.

The 40-member Independent International Scientific Panel on AI held its first plenary meeting virtually, three weeks after the General Assembly appointed its members. Members elected Yoshua Bengio and Maria Ressa as co-chairs. The Secretary-General asked members to produce independent, multidisciplinary and evidence-based assessments of AI opportunities, risks and impacts. Under its terms of reference, the Panel may issue an annual report, thematic briefs and up to two annual updates to the General Assembly.

The Scientific Panel held its first in-person meeting in Madrid. Members worked on the evidence base for the Panel’s first assessment of AI capabilities, opportunities, risks and impacts. The meeting was an important drafting milestone ahead of the inaugural Global Dialogue.

ODET launched the AI Governance for Humanity Lab in Valencia, Spain. The Lab is intended to support applied policy research, experimentation, capacity-building and inclusive multi-stakeholder dialogue. It advances recommendations from the Advisory Body’s final report and complements the Scientific Panel and Global Dialogue.

The Scientific Panel released the Preliminary Report of the Independent International Scientific Panel on AI: Evidence-based assessment of opportunities, risks and impacts of AI. The report states that AI capabilities are advancing faster than safeguards and governance capacity and highlights unequal access to compute, data, skills and benefits. The Secretary-General launched the report with co-chairs Yoshua Bengio and Maria Ressa. The official publication is titled a Preliminary Report, although some event material referred more generally to the Panel’s first report or annual report.

The inaugural Global Dialogue met at the Palexpo International Convention Centre in Geneva, co-chaired by El Salvador and Estonia. Governments, technology companies, civil society, academia and the technical community discussed the seven themes set out in resolution 79/325: safe, secure and trustworthy AI; capacity-building; social, economic, ethical, cultural, linguistic and technical implications; interoperability of governance approaches; human rights; transparency, accountability and human oversight; and open-source software, open data and open models. Six months of consultations preceded it, drawing more than 1,500 written submissions from every UN regional group. The Scientific Panel presented its Preliminary Report. The Dialogue was a separate UN event from the AI for Good Global Summit, although both were held at Palexpo and their dates overlapped.

Ambassador Egriselda López of El Salvador and Ambassador Rein Tammsaar of Estonia co-chaired the 2026 Dialogue. They guided the six months of consultations, issued the note on themes and structure, and steered the formal sessions. By design each Dialogue closes with a co-chairs’ summary rather than a negotiated outcome, so that every delegation participates on equal footing. The summary of the Geneva session had not yet been posted at a stable official URL as of 28 July 2026.

The 2026 summit took place at Palexpo in Geneva from 7 to 10 July. It overlapped with the second day of the Global Dialogue but was a separately registered event. Its programme addressed standards, skills, policy, robotics and AI applications for global challenges.

Mandated by resolution 79/239, the Secretary-General submitted a consolidated summary of the views received from Member States and observer States on artificial intelligence in the military domain, with specific focus on areas other than lethal autonomous weapons systems. The report catalogues opportunities, challenges and existing or emerging normative proposals, and annexes the submissions themselves. It was issued to the eightieth session and reissued for technical reasons on 4 February 2026.

The Scientific Panel is mandated to produce an annual evidence-based report on AI opportunities, risks and impacts for presentation at the Global Dialogue. It may also issue thematic briefs, establish working groups, consult external experts and update the General Assembly up to twice annually. Its mandate excludes military AI.

The Global Digital Compact mandates a high-level review during the General Assembly’s eighty-second session. The review is intended to assess implementation, identify gaps and consider further action. Exact dates, preparatory reports and stakeholder modalities remain to be decided.

The second Global Dialogue is confirmed for May 2027 in New York. It is expected to receive the Panel’s next annual assessment and continue discussion of safe and trustworthy AI, capacity-building, interoperability and inclusion. Exact dates, programme, venue and co-chairs had not been publicly confirmed as of 28 July 2026.