The Global Digital Compact contains the first truly universal commitments on AI governance, adopted by consensus.
Hand-checked · A/RES/79/1, Annex I
The General Assembly established the Independent International Scientific Panel on AI and the Global Dialogue on AI Governance, without a vote.
Hand-checked · A/RES/79/325
The Assembly adopted its first resolution devoted broadly to artificial intelligence in March 2024, by consensus.
Hand-checked · A/RES/78/265
The UNESCO Recommendation on the Ethics of Artificial Intelligence was the first global normative instrument on the subject.
Hand-checked · 41 C/Resolution 15
A high-level review of the Global Digital Compact is mandated for the Assembly’s eighty-second session.
Hand-checked · A/RES/79/1
Safe line: “The second Global Dialogue in May 2027 is the next mandated global stocktaking point for the UN AI governance architecture.”
Safe line: “The Global Dialogue is the Assembly’s stocktaking point on AI governance.” Put May 2027 in the PGA planning grid; ask early for a modalities checklist.
Tell capital: if you staff one AI meeting, make it New York, May 2027; use the Global Dialogue roadmap and FAQ until formal details post.
Safe line: “The Global Digital Compact itself mandates a high-level review to assess implementation, identify gaps and consider further action.” Cite A/RES/79/1.
Cite A/RES/79/1: “The eighty-second-session review is the formal accountability moment for implementation, gaps and further action.”
Cite A/RES/79/1 when requesting capital inputs. Use a three-part checklist: implementation, gaps, further action.
Cite A/RES/79/325, and pull review-ready language from the Geneva programme, webcasts and the Panel’s Preliminary Report.
Cite resolution 79/325 and the Dialogue programme to show these themes already structure UN AI-governance discussions; use the FAQ if someone conflates it with AI for Good.
Cite resolution 79/325 and the inaugural Geneva Dialogue as proof development priorities belong in AI-governance debates; use the seven themes as your meeting checklist.
Safe line: “The first universal UN forum mandated specifically for AI governance is now operational.” Cite A/RES/79/325.
Cite A/RES/79/325 and the Geneva programme to keep rights, accountability and human oversight on the agenda, not blurred with the overlapping AI for Good Summit.
Cite A/RES/79/325. Safe line: “The Assembly’s universal AI forum is operational.” Also note: it was separate from the AI for Good Global Summit.
Cite A/RES/79/325 for the seven themes; use the Geneva programme and WebTV record to brief colleagues on who is shaping the conversation.
Tell capital: the 6–7 July 2026 Dialogue covered all seven themes in A/RES/79/325 and heard the Panel’s Preliminary Report; cite A/RES/79/325 and the webcast.
Cite the full title and quote this line: AI capabilities are advancing faster than safeguards and governance capacity.
Cite the Preliminary Report by title in SDG and UNCT notes, especially its line on unequal access to compute, data, skills and benefits.
Safe line: “The Preliminary Report shows AI is advancing faster than safeguards.” Cite the official Preliminary Report page.
Cite the Panel members page and FAQ: appointed by the General Assembly; members serve in their personal capacity, not for governments or institutions.
Safe line: “The General Assembly appointed a 40-member independent scientific panel on AI, serving in their personal capacity across all five UN regions.”
Safe line: “The General Assembly appointed the members; the members elected the co-chairs and serve in their personal capacity.”
Cite A/RES/79/325 for the baseline: non-military scope, 40-member independent panel, geographical balance, annual reporting, inclusive annual dialogue.
General Assembly establishes the Scientific Panel and Global Dialogue →
Cite A/RES/79/325 for any proposal needing a UN-backed AI governance reference point, and use A/79/L.118 and A/79/PV.89 to trace text and adoption.
General Assembly establishes the Scientific Panel and Global Dialogue →
Cite A/RES/79/325 to back inclusive AI-governance support, and use the annual Dialogue to raise development priorities and the AI divide.
General Assembly establishes the Scientific Panel and Global Dialogue →
Cite A/RES/79/325. Provenance: A/79/L.118; adoption record A/79/PV.89.
General Assembly establishes the Scientific Panel and Global Dialogue →
Cite A/RES/79/325 for terms of reference and modalities; for process history, use A/79/L.118 and A/79/PV.89.
General Assembly establishes the Scientific Panel and Global Dialogue →
Cite A/RES/79/325. Safe line: “The General Assembly established the UN’s standing AI science-and-dialogue mechanisms by consensus.”
General Assembly establishes the Scientific Panel and Global Dialogue →
Cite A/RES/79/325 in briefs to draw the line: universal UN AI assessment and dialogue exist, but their mandate stops at the non-military domain.
General Assembly establishes the Scientific Panel and Global Dialogue →
Cite A/RES/79/325 for agreed terms; if capital asks scope, the clean line is that the Panel’s mandate is limited to the non-military domain.
General Assembly establishes the Scientific Panel and Global Dialogue →
Start every standards scan here, using the AI Standards Exchange database to show alignment with Global Digital Compact commitments on interoperable standards and international cooperation.
In pitches and scopes, anchor your work to ODET’s remit: AI governance, data governance, digital inclusion, digital public infrastructure, connectivity, and emerging technologies.
Office for Digital and Emerging Technologies begins operation →
Use the AI Standards Exchange database to source standards and technical publications for concept notes, procurement specs and internal interoperability checklists.
Cite ODET as the central Secretariat office for GDC and AI-governance follow-up in coordination asks, architecture discussions and interoffice role-setting.
Office for Digital and Emerging Technologies begins operation →
Cite A/RES/70/125 for the standing digital-development frame, and use the WSIS action lines as a crosswalk for AI capacity-building and inclusion.
Point teams to the ITU AI Standards Exchange database before speeches or briefing notes. Safe line: “ITU’s AI Standards Exchange helps navigate a fragmented international AI standards landscape.”
Safe line: “ODET is the Secretariat hub for Global Digital Compact follow-up and AI governance coordination.” Bring ODET in early on speeches, briefs and meeting prep.
Office for Digital and Emerging Technologies begins operation →
Cite A/RES/70/125 in PGA lines on continuity: AI governance should build through existing UN digital frameworks, not float above them.
Cite A/RES/70/125, and ask capital whether your AI lines also cover connectivity, inclusion and Internet-governance language.
Cite A/RES/79/239 to insist scrutiny covers full military AI functions, not only weapons release decisions; submit rights-focused views for the Secretary-General’s report.
Cite A/RES/79/239 to raise military AI in mission technology, early warning or Security Council-facing analysis beyond the LAWS lane.
Tell capital this passed by recorded vote, 159-2-5, on 24 December 2024; cite A/RES/79/239 as the GA entry point on military AI.
Cite S/PV.9821. Safe line: the Secretary-General warned that AI is outpacing the world’s ability to govern it and urged Member States to agree international guard-rails.
Security Council high-level debate on AI and international peace and security →
Cite S/PV.9821 and the Secretary-General’s line that AI is outpacing the world’s ability to govern it to back early-warning, mission-technology and protection discussions.
Security Council high-level debate on AI and international peace and security →
Cite S/PV.9821: the Council treated AI as a peace and security issue. Tell capital this may need political, legal and tech coordination.
Security Council high-level debate on AI and international peace and security →
Cite A/RES/79/1, Annex I, then check your prototype against its headings: data governance, digital public infrastructure, trust and safety, rights online, AI governance.
Cite A/RES/79/1, Annex I in concept notes and standards discussions to justify interoperability, inclusion, rights, and data-governance requirements together.
Cite A/RES/79/1, Annex I in notes and funding pitches: the intergovernmental mandate for the UN’s digital and AI governance architecture.
Cite A/RES/79/1, Annex I. Safe line: “The Global Digital Compact is the UN’s consensus mandate on digital and AI governance.”
Cite A/RES/79/1, Annex I. Useful line: governments adopted commitments on human rights online, trust and safety, and AI governance by consensus.
Cite A/RES/79/1, Annex I. Safe line: “The UN’s principal intergovernmental mandate for digital and AI governance.”
Cite A/RES/79/1, Annex I. Handy line: the principal intergovernmental mandate for the UN’s current digital and AI governance architecture.
Cite A/RES/79/1, Annex I. Safe line: the intergovernmental mandate for the UN’s current digital and AI governance architecture.
Cite Governing AI for Humanity for standards-sharing, capacity-building, data-governance or funding asks; use its seven measures as your concept-note structure.
Cite Governing AI for Humanity for a UN-rooted menu of technical options and a line linking AI governance to the UN Charter, international law, human rights and sustainable development.
Cite Governing AI for Humanity on capacity-building and finance, and use its seven proposals as a checklist in development-policy discussions.
Cite Governing AI for Humanity: UN AI governance should be grounded in the UN Charter, international law, human rights and sustainable development.
Quote that baseline when challenging surveillance-heavy or rights-thin AI proposals; attach the 19 September 2024 final report link.
Safe PGA line: “Member States now have a concrete menu for UN-grounded AI governance, rooted in the Charter, international law, human rights and sustainable development.”
Cite Governing AI for Humanity for a Charter-, international law- and human rights-based frame in briefings on AI in conflict or mission technology.
Cite Governing AI for Humanity and use the seven-part list as your checklist for capital: what is in, out, or still open.
Cite A/RES/78/311 or A/78/L.86 in concept notes to justify capacity-building components: financing, knowledge-sharing, data, skills and institutional capacity are named.
Cite A/RES/78/311 in programme notes and funding pitches; use its list as a checklist for what an AI capacity-building package should cover.
Cite A/RES/78/311, adopted without a vote on 1 July 2024; safe line: "The General Assembly has already set equitable AI capacity-building as a shared priority."
Cite A/RES/78/311; if process matters, note it was China-led, issued as A/78/L.86, and adopted without a vote on 1 July 2024.
Cite A/RES/78/265 and use its themes as your build checklist: international law, human rights, privacy, safety, inclusion, sustainable development, digital and development divides.
Cite A/RES/78/265; for the draft and adoption record, use A/78/L.49 and A/78/PV.63.
Use A/RES/78/265 as agreed language linking AI and sustainable development; consensus adoption makes it a safe cross-system reference.
Use A/RES/78/265. Safe line: “The General Assembly has already adopted by consensus a broad baseline for safe, secure and trustworthy AI systems for sustainable development.”
Cite A/RES/78/265 and, if needed, A/78/PV.63 to anchor interventions on international law, human rights, privacy, safety and inclusion.
Cite A/RES/78/265, adopted on 21 March 2024, and say: this is the Assembly’s first universal consensus on broad AI governance.
Cite A/RES/78/265 as the first universal GA consensus on broad AI governance, adopted without a vote on 21 March 2024.
Cite A/RES/78/265: adopted by consensus, without a vote, from A/78/L.49 on 21 March 2024.
Use OCHA’s public data-responsibility framework as your checklist, and point reviewers to the Centre for Humanitarian Data and OCHA anticipatory action pages.
OCHA data responsibility, predictive analytics and anticipatory action →
Use “from the digital divide to the AI divide” in planning and financing talks, backed by A/RES/78/311 and cited UN reports.
Cite A/RES/79/1 for capacity-building, financing and AI-divide asks, then pair it with A/RES/79/325 when you need the implementation track, not just the promise.
Apply OCHA’s public data-responsibility framework as a checklist in programme design and financing talks on AI, forecasts and pre-agreed triggers.
OCHA data responsibility, predictive analytics and anticipatory action →
Safe line: “Closing the AI divide is a governance task, not an afterthought.” Cite A/RES/78/311.
Cite A/RES/79/1 for the commitments and A/RES/79/325 for first implementation. Safe line: “The Compact now has a live review cycle, not just a launch moment.”
Use: closing the AI divide is a governance task, not a side programme. Anchor it in A/RES/78/311 and the cited UN reports.
Cite A/RES/79/1: the Assembly is where AI commitments, operational mechanisms and review meet. Safe line: “the General Assembly is where AI commitments return for accountability.”
Apply OCHA’s public data-responsibility framework as a checklist for early-warning, forecasting and mission-support tools that use humanitarian data.
OCHA data responsibility, predictive analytics and anticipatory action →
Brief capital with A/RES/79/1 → A/RES/79/325 → GA eighty-second-session review: the clearest citation trail on implementation and where process space remains open.
Cite A/RES/78/241 to argue that LAWS now warrants Assembly-wide scrutiny, and submit rights-focused views when the Secretary-General seeks stakeholder input.
General Assembly resolution on lethal autonomous weapons systems →
Cite A/RES/78/241, adopted on 22 December 2023 by recorded vote. Safe line: “The Assembly opened a universal track while recognizing the CCW’s continuing work.”
General Assembly resolution on lethal autonomous weapons systems →
Cite A/RES/78/241: the Assembly opened a universal track on lethal autonomous weapons systems while recognizing continuing CCW work.
General Assembly resolution on lethal autonomous weapons systems →
Cite A/RES/78/241 to brief capital: the Assembly has opened a parallel universal LAWS track; submit national views early for the Secretary-General’s report.
General Assembly resolution on lethal autonomous weapons systems →
Cite Governing AI for Humanity as the first consolidated UN expert proposal for inclusive international AI governance, and use the functions as your checklist.
Cite Governing AI for Humanity and quote “accountability” when you argue AI governance cannot stop at standards and capacity-building.
Cite Governing AI for Humanity as the first consolidated UN expert proposal: scientific assessment, risk monitoring, policy interoperability, capacity-building, accountability.
Cite the Secretary-General’s 26 October 2023 launch and the UN AI Advisory Body pages to show you are aligning early with emerging UN AI-governance expectations.
Cite the Secretary-General’s 26 October 2023 launch and the official AI Advisory Body pages to separate expert advice from agreed UN policy in briefs, guidance and adoption reviews.
Cite the 26 October 2023 launch and use a three-part test in proposals: governance, shared understanding, and AI for the SDGs.
Safe line: “On 26 October 2023, the Secretary-General launched a 39-member High-level Advisory Body on AI from 33 countries.”
Cite the 26 October 2023 launch and the UN AI Advisory Body pages to justify rights-based input now.
Safe PGA line: “The Secretary-General convened expert advice; Member States decide what becomes intergovernmental mandate.”
Need the origin point for current UN AI-governance ideas? Cite the Secretary-General’s launch of the High-level Advisory Body on AI on 26 October 2023.
Tell capital this is advisory, not negotiated text. Cite the Secretary-General’s 26 October 2023 launch and mandate when tracing later UN AI proposals.
Safe line: “The Secretary-General and ICRC President have called for a legally binding instrument on autonomous weapons, to be concluded by 2026.”
Secretary-General and ICRC President call for a legally binding instrument on autonomous weapons →
Quote the 5 October 2023 joint call: prohibit unpredictable systems, prohibit systems applying force against persons, regulate others through limits and meaningful human control.
Secretary-General and ICRC President call for a legally binding instrument on autonomous weapons →
Quote the 5 October 2023 joint call in mission advice, mandate inputs, and Council lines when you need a concise standard.
Secretary-General and ICRC President call for a legally binding instrument on autonomous weapons →
Cite ILO Working Paper 96: clerical work has the highest exposure, and job transformation is more likely than full job elimination.
Cite ILO Working Paper 96 (21 August 2023). Safe line: “Job transformation is more likely than full elimination, with the highest exposure in clerical work and clear gender differences.”
Cite ILO Working Paper 96 to press for due diligence on occupation- and sex-based impacts, not just job-loss claims.
Cite S/PV.9381 and check your design for disinformation, cyberattack and weapons-risk exposure before review.
Cite S/PV.9381 and 18 July 2023 to show AI reached the Council agenda; use it to justify risk, safety and policy-coordination work.
Cite S/PV.9381: AI can support development, humanitarian action and peace efforts, but also amplify disinformation, cyberattacks and weapons risks.
Cite S/PV.9381. Safe line: “The Security Council has already treated AI as a matter of international peace and security.”
Cite S/PV.9381: the Council heard AI framed as carrying development promise and conflict risk, useful when pressing safeguards and accountability in security debates.
Cite S/PV.9381 and use: security risks are real, but universal AI governance belongs in member-state forums.
Cite S/PV.9381 to justify AI in mission analysis, briefings or mandate language: AI can aid peace efforts but also amplify disinformation, cyberattacks and weapons risks.
Cite S/PV.9381 and 18 July 2023 to show the Council formally took up AI; tell capital this is not only a digital-policy track.
Use it to scout collaborators and test whether your prototype can survive both a demo and a standards conversation.
Use “AI for Good” as a convening hook when you need one frame linking SDG delivery, accessibility, sector policy and standards communities.
Safe line: "AI for Good links fast-moving AI with the UN’s development and standards communities." Proof point: the 6–7 July 2023 Geneva summit.
Use those public examples as precedent and design to the target entity’s review process; there is no single UN-wide compliance checklist here.
Apply Updated UN Guidance on the Use of AI Tools at Work as your checklist; cite the CEB model policy framework and operational-use report.
Use WFP SKAI, UNHCR Project Jetson, the Centre for Humanitarian Data and OCHA anticipatory-action pages as proof this work is operational, not declaratory.
Cite Updated UN Guidance on the Use of AI Tools at Work (5 March 2025) in rollout notes and staff checklists.
Safe line: “The UN already has staff guidance on AI tools at work.” Cite: Updated UN Guidance on the Use of AI Tools at Work.
Cite WFP’s SKAI, UNHCR’s Project Jetson, OCHA’s Centre for Humanitarian Data, and anticipatory action work as concrete UN examples.
Use 30 November 2022 as the before/after line in concept notes or briefings to explain why your prototype now gets leadership scrutiny.
Use “the event that transformed the pace and prominence of everything the UN did next”; cite 30 November 2022.
Use 30 November 2022 as the clean before/after line in briefs: before, specialist topic; after, leadership priority.
Safe line: “ChatGPT made generative AI a mainstream public phenomenon and pushed AI to the top of global agendas.”
Use 30 November 2022 as your before/after line in briefings: after ChatGPT’s release, AI moved to the top of multilateral agendas worldwide.
Safe PGA line: “ChatGPT’s launch turned AI from a specialist topic into a leadership priority worldwide.” Scene-setting only, not a UN mandate.
Use 30 November 2022 as the before/after line in briefs: after ChatGPT’s launch, AI moved to the top of political, economic and multilateral agendas.
Tell capital: this is the trigger for the UN’s accelerated AI work. In briefings, anchor the timeline with “ChatGPT launched on 30 November 2022.”
Cite CEB/2022/2/Add.1 in concept notes and map your project to the ten principles before review: safety, privacy, oversight, transparency, accountability, inclusion, and more.
Principles for the Ethical Use of AI in the United Nations System →
Cite CEB/2022/2/Add.1 in concept notes, model reviews, and governance checklists; map proposals against all ten principles before approval.
Principles for the Ethical Use of AI in the United Nations System →
Cite CEB/2022/2/Add.1 and test any AI proposal against the listed principles, especially proportionality, privacy, transparency, explainability, responsibility, and accountability.
Principles for the Ethical Use of AI in the United Nations System →
Cite 41 C/Resolution 15 and map your system to privacy, data governance, transparency, accountability and human rights before review does it for you.
UNESCO Recommendation on the Ethics of Artificial Intelligence →
Cite 41 C/Resolution 15, adopted 23 November 2021, then check your project against its domains: transparency, accountability, privacy, data governance, labour, health, environment.
UNESCO Recommendation on the Ethics of Artificial Intelligence →
Cite 41 C/Resolution 15 as the first global normative instrument on AI ethics; use its topic list as a design checklist.
UNESCO Recommendation on the Ethics of Artificial Intelligence →
Cite 41 C/Resolution 15. Safe line: “UNESCO’s Recommendation on the Ethics of Artificial Intelligence is the first global normative instrument on AI ethics, adopted by consensus.”
UNESCO Recommendation on the Ethics of Artificial Intelligence →
Cite 41 C/Resolution 15 and UNESCO’s line: the first global normative instrument on AI ethics.
UNESCO Recommendation on the Ethics of Artificial Intelligence →
Cite 41 C/Resolution 15: “A consensus normative reference on AI ethics exists; the Assembly must steer follow-up, review and political direction.”
UNESCO Recommendation on the Ethics of Artificial Intelligence →
Cite UNESCO’s Recommendation on the Ethics of Artificial Intelligence, 41 C/Resolution 15, and test proposals against rights, privacy, accountability and transparency.
UNESCO Recommendation on the Ethics of Artificial Intelligence →
Cite 41 C/Resolution 15. Handy line for capital: this is the first global normative instrument on AI ethics, adopted by consensus.
UNESCO Recommendation on the Ethics of Artificial Intelligence →
Anchor proposals in “UN 2.0” and the “Quintet of Change,” and cite the official pages to show your AI work fits the UN’s modernization frame.
Use: “AI is part of digital and data transformation, not an isolated reform stream,” and point colleagues to the UN 2.0 site and network.
Safe line: “We are building AI capability through UN 2.0’s Quintet of Change to make programmes more adaptive, evidence-based and future-oriented.” Link: un.org/two-zero/en.
Cite A/74/821 when you need the Secretary-General’s early, system-wide basis for AI cooperation inside broader digital-cooperation work.
Cite A/74/821. Safe line: “The Roadmap for Digital Cooperation put stronger international AI cooperation on the UN policy track.”
Cite A/74/821 to show the Secretariat’s policy line started early; safe line: “GA authority begins when Member States turn proposals into mandates.”
Use the Secretary-General’s Data Strategy page as your checklist: show how your prototype improves decision-making, programme delivery or organizational management and uses governed, reliable data.
Use the Secretary-General’s Data Strategy page to anchor business cases and reviews: reliable, governed data is a prerequisite for responsible institutional AI.
Safe line: “Responsible AI in the Secretariat starts with responsible data.” In briefs, tie AI proposals to the strategy’s governance, quality and skills foundations.
Use Y.3172, Y.3173 and Y.3174 as checks for network integration, intelligence-level evaluation, and machine-learning data handling, then cite the relevant ITU-T Recommendation in your spec.
Cite Y.3172, Y.3173 and Y.3174 for integrating ML into networks, evaluating intelligence levels, and handling ML data.
Cite Y.3172 on integrating machine learning into future networks, Y.3173 on evaluating intelligence levels, and Y.3174 on data handling for machine learning.
Show up to AI for Good when you need partners, use cases or standards signals; cite the first summit: 7–9 June 2017, ITU.
Use “AI for Good” in concept notes and partnership outreach as an established UN platform for practical AI and the SDGs.
Safe line: "AI for Good is the UN system’s longest-running global platform on AI applications, standards and partnerships."
Cite the agreed guiding principles and the CCW/GGE.1 series when you push accountability for weapon-use decisions.
Cite the CCW/GGE.1 series and the guiding principles in briefs on LAWS, mission technology, or protection risks.
Cite the CCW/GGE.1 series when reporting to capital. Safe line: the CCW remains the principal specialized diplomatic process on autonomous weapons.
No line matches that. Clear a filter, or try the words you would actually say: “capacity building”, “consensus”, “review”.