Opening Ceremony

8 intervenants
Résumé

Résumé

La discussion a ouvert le Dialogue mondial des Nations Unies sur la gouvernance de l’IA en tant que premier forum mondial destiné à lancer un processus plus large et universel réunissant les États, les organisations internationales, la société civile, le monde universitaire et les experts techniques autour de la gouvernance de l’IA . Les coprésidents, l’ambassadrice Egriselda López d’El Salvador et l’ambassadeur Rein Tammsaar d’Estonie, ont souligné que le dialogue devait être inclusif, fondé sur des données probantes et orienté vers la coopération pratique afin que l’IA contribue au développement et profite à tous les pays, sans en laisser certains de côté . António Guterres a affirmé que l’IA progresse à une « vitesse effrénée » et transforme déjà les économies, le travail, les élections et la sécurité plus rapidement que les institutions ne peuvent le gérer, rendant la situation actuelle intenable sans élaboration de plan commun . Il a déclaré que le premier rapport du Groupe scientifique international indépendant sur l'intelligence artificielle fournirait une base factuelle commune et il a mis en garde contre trois grands dangers : la rapidité du déploiement de l’IA, la concentration du pouvoir entre les mains de quelques entreprises et pays, et l’érosion de la vérité via la désinformation facilitée par les machines . Dans le même temps, il a souligné le potentiel positif de l’IA pour la santé, l’éducation et l’agriculture, estimant que, si ses bénéfices étaient largement partagés, l'IA pourrait accélérer le développement et même devenir un « grand facteur d'égalisation » . António Guterres a ensuite exposé quatre priorités pour l’action mondiale : des normes de sécurité communes, des lignes rouges ancrées dans les droits humains, un renforcement des capacités pour les pays en développement, et une plus grande transparence sur l’empreinte environnementale de l’IA . Il a également appelé à un engagement pour la sécurité des enfants face à l’IA, et a demandé un fonds et un réseau mondial pour les capacités en matière d’IA, ainsi qu’une interdiction en droit international des armes létales autonomes, qu’il a qualifiées de « robots tueurs » . Annalena Baerbock a renforcé le sentiment d’urgence, déclarant que le rythme de développement et la complexité de l’IA signifient qu’elle ne peut être gouvernée par quelques acteurs seulement et qu’elle doit être encadrée par l’ONU avec la participation de tous les pays et de toutes les parties prenantes . D’autres intervenants se sont concentrés sur la mise en œuvre et le soutien institutionnel. Doreen Bogdan-Martin a déclaré que l’ONU avait avancé plus rapidement sur l’IA qu’elle ne l’avait fait pour l’Internet naissant et a décrit la Geneva Digital Week comme un effort coordonné reliant ce dialogue, le panel scientifique, le AI for Good Summit et le Forum du SMSI . Khaled El-Enany a mis en avant le cadre éthique de l’UNESCO, le soutien apporté à plus de 80 pays, la formation de plus de 50 000 responsables publics et acteurs judiciaires, ainsi que les nouveaux travaux sur la protection des enfants et des jeunes à l’ère de l’IA . Amandeep Singh-Gill a souligné l’ampleur de la participation, la nécessité d’une inclusion soutenue par des capacités réelles, et l’intention de poursuivre le processus avec un deuxième cycle à New York l’année prochaine . Dans l’ensemble, l’ouverture a établi un large consensus selon lequel la gouvernance de l’IA doit être inclusive, internationale, pratique et gérée de manière urgente pour que l’IA soit sûre, équitable et bénéfique pour toute l’humanité .

Points clés

- L’objectif central du dialogue est de lancer le premier processus onusien mondial, universel et inclusif sur la gouvernance de l’IA, réunissant tous les pays et toutes les parties prenantes afin de partager des données probantes, des perspectives et des expériences, plutôt que de laisser la politique de l’IA à un petit nombre d’acteurs. Les intervenants ont souligné à plusieurs reprises que chaque pays devait avoir sa place à la table et que la participation devait inclure les gouvernements, les organisations internationales, la société civile, le monde universitaire, les entreprises et les experts techniques. - La nécessité d’équilibrer les vastes opportunités offertes par l’IA avec les risques sérieux a été un thème majeur. Les participants ont mis en avant le potentiel de l’IA pour faire progresser le développement, la santé, l’éducation, l’agriculture et les Objectifs de développement durables (ODD), tout en avertissant que l'IA pouvait remodeler les marchés du travail, influencer les élections, fragiliser la sécurité et altérer le tissu social si elle n’est pas encadrée. - Plusieurs intervenants ont estimé que la gouvernance de l’IA doit être fondée sur les droits humains, la sécurité, la responsabilité et la confiance. Pour ce faire, des appels à des références internationales communes pour les tests et l’évaluation des risques, des appels au maintien de la responsabilité humaine dans les décisions à forts enjeux, et à la garantie que l’IA reste éthique, sûre et digne de confiance sont nécessaires. - Un autre point clé de discussion a été l’inégalité en matière de pouvoir et de capacités dans l’IA, en particulier le risque que les pays en développement soient exclus. António Guterres a averti que la puissance de calcul, les données et les talents sont concentrés dans une poignée d’entreprises et de pays, tandis que plusieurs intervenants ont souligné la nécessité de compétences, d’infrastructures, de financements, de capacités institutionnelles et de partenariats afin que tous les pays puissent participer et bénéficier de l'IA. - La sécurité des enfants, les utilisations dommageables et d’autres lignes rouges sont apparues comme une priorité de gouvernance particulièrement urgente. António Guterres a appelé à un engagement pour la sécurité des enfants face à l’IA, en citant des préjudices tels que la manipulation, l’automutilation et les images d’abus, tandis qu'Annalena Baerbock a mis en lumière la mauvaise utilisation de l’IA avec des deepfakes et des violences genrées. António Guterres a également tracé une ligne rouge ferme contre les armes létales autonomes, les qualifiant de « robots tueurs » inacceptables sur le plan moral et nécessitant une interdiction juridique internationale.

L’objectif général de la discussion était d’ouvrir le Dialogue mondial des Nations Unies sur la gouvernance de l’IA et d’en définir l’ordre du jour : créer un processus multilatéral inclusif capable de traduire les données scientifiques et la consultation mondiale en coopération internationale pratique, en normes communes et en actions futures sur la gouvernance de l’IA. Les intervenants ont présenté la réunion comme le point de départ d’un processus continu qui se poursuivra au-delà de Genève et contribuera à garantir que l’IA soit plus sûre, plus équitable, plus accessible et bénéfique pour toute l’humanité.

Le ton général était formel, urgent et coopératif. La discussion a commencé de manière cérémonielle et diplomatique, avec des remerciements et une insistance sur l’inclusion et le partenariat. Le ton est ensuite devenu plus prudent et plus ferme pendant les remarques d’António Guterres, en particulier lorsqu’il a été question de changement technologique effréné, de concentration du pouvoir, de désinformation, de préjudices causés aux enfants et d’armes létales autonomes. Vers la fin, le ton est devenu plus optimisme et pragmatique, dénotant un sens de responsabilité collective, et soulignant que la coopération multilatérale peut encore orienter l’IA vers le bien public si l’action se poursuit.

Intervenants

- Doreen Bogdan-Martin - Secrétaire générale de l’Union internationale des télécommunications (UIT) ; co-coordinatrice du Secrétariat conjoint pour le Dialogue sur l’IA [S1] - António Guterres - Secrétaire général des Nations Unies [S4] - Announcer - Annonceur de la cérémonie d’ouverture - S.E. l’Ambassadeur Rein Tammsaar, Estonie - Coprésident du Dialogue mondial sur la gouvernance de l’IA ; Ambassadeur d’Estonie ; ancien ambassadeur numérique de l’Estonie [S11] - Khaled El-Enany - Directeur général de l’UNESCO ; co-coordinateur du Secrétariat conjoint pour le Dialogue sur l’IA [S13] - S.E. l’Ambassadrice Egriselda López, El Salvador - Coprésidente du Dialogue mondial sur la gouvernance de l’IA ; Ambassadrice d’El Salvador - Annalena Baerbock - Présidente de l’Assemblée générale des Nations Unies [S18] - Amandeep Singh-Gill - Secrétaire général adjoint des Nations Unies et envoyé spécial pour les technologies numériques et émergentes, ODET [S20][S21]

Intervenants supplémentaires : - Aucun

Intervenants
HA
H.E. Ambassador Egriselda López, El Salvador
93 wpm · 6 min
HA
H.E. Ambassador Rein Tammsaar, Estonia
97 wpm · 6 min
AB
Annalena Baerbock, President of the General Assembly
108 wpm · 10 min
AS
Amandeep Singh-Gill, USG and Special Envoy for Digital and Emerging Technologies, ODET
147 wpm · 3 min
DB
Doreen Bogdan-Martin
106 wpm · 4 min
KE
Khaled El-Enany, Director-General, UNESCO
136 wpm · 3 min
AG
António Guterres
111 wpm · 17 min
A
Announcer
84 wpm · 53 s

La session a officiellement ouvert le premier Dialogue mondial des Nations Unies sur la gouvernance de l’IA et l’a présenté comme le début d’un processus multilatéral plus long plutôt que comme une conférence isolée . L’Ambassadrice Egriselda López d’El Salvador a présenté la réunion comme le lancement d’un espace large et véritablement universel où différentes expériences, capacités et catégories d’acteurs pourraient converger, en s’appuyant sur un mandat de l’Assemblée générale ainsi que sur des années d’innovation, de production de connaissances et de pratiques partagées entre les États, les organisations internationales, la société civile, le monde universitaire et les experts techniques . Les coprésidents et le secrétariat ont été remerciés d'avoir contribué à transformer le mandat des États membres en dialogue qui s’ouvre aujourd’hui . Egriselda López a également souligné que Genève devait être considérée non seulement comme un lieu d’arrivée, mais aussi comme un point de départ pour un travail continu . Les intervenants ont inscrit le Dialogue dans le mandat créé par le Pacte pour l’avenir et le Pacte numérique mondial, et plusieurs d’entre eux ont insisté sur le fait qu’il devait se poursuivre au-delà de cette réunion inaugurale .

Egriselda López et Rein Tammsaar pour l'Estonie ont exposé le cadre de base de la réunion par les coprésidents autour de l’inclusion, des capacités et de l’instauration de la confiance . Egriselda López a soutenu que la valeur du dialogue ne réside pas seulement dans l’échange de positions formelles, mais aussi dans le partage d’expériences, l’apprentissage mutuel et le renforcement de la coopération internationale face à une technologie qui évolue rapidement et exige donc des réponses agiles et inclusives . Elle a également souligné qu’une participation effective signifie plus qu’une simple présence physique : les pays ont besoin de compétences, d’infrastructures, de talents, de financements, d’institutions, d’alliances et de partenariats s’ils veulent façonner l’IA et en tirer parti sur un pied d’égalité . Rein Tammsaar a renforcé ce point en observant que l’IA touche déjà tous les pays, quel que soit leur niveau de développement technologique, de sorte que la conversation sur la gouvernance doit inclure toutes les régions, tous les niveaux de développement et toutes les parties prenantes pertinentes, sans laisser personne de côté . Il a décrit la tâche des coprésidents comme celle de créer une plateforme universelle où des perspectives diverses, des connaissances scientifiques, leunjugement politique et une délibération technologique pourraient se rejoindre afin de bâtir une confiance et une compréhension commune .

Ce cadre d’ouverture était lié à une deuxième idée centrale : la gouvernance de l’IA doit équilibrer les opportunités et les responsabilités . Egriselda López a explicitement appelé à un équilibre entre innovation et responsabilité, opportunité et risque, coopération internationale et réalités nationales, ainsi qu’entre la prise de décision mondiale et les besoins concrets de chaque pays . Rein Tammsaar a de même soutenu que les nombreuses possibilités offertes par l’IA ne peuvent être pleinement réalisées que si les systèmes sont sûrs, ancrés dans le droit international et les droits humains, et façonnés par les personnes elles-mêmes plutôt que leur étant imposés . S’appuyant sur l’expérience de l’Estonie, il a déclaré que les outils numériques peuvent améliorer considérablement la vie quotidienne lorsqu’ils sont centrés sur l’humain, dignes de confiance et soutenus par des institutions publiques solides, tout en avertissant que la technologie, à elle seule, ne suffit jamais . Il a également précisé que l’objectif des deux jours suivants n’était pas d’effacer les différences, mais de mieux les comprendre, d’identifier les domaines de convergence, de recueillir les bonnes pratiques et de poser les fondations d’une coopération future .

António Guterres a ensuite déplacé la discussion en avertissant sur l’urgence . Il a déclaré que l’IA progresse à une « vitesse folle » et qu’elle est déployée plus rapidement que quoi que ce soit, ses concepteurs ne pouvant véritablement pas la suivre, alors qu’elle peut remodeler les économies, transformer les marchés du travail, avoir une influence sur les élections et modifier l’équilibre de la sécurité . La formulation la plus forte d'António Guterres a été qu’« une expérience est menée sur notre planète et sur les sociétés sans plan et sans consentement », ce qu’il a qualifié à la fois d’insoutenable et d’inacceptable . Il a présenté la question centrale de savoir si l’humanité façonnera la transformation de l’IA ou laissera au contraire la technologie façonner l’humanité. Il a déclaré que ce Dialogue devait apporter le début d’une réponse parce que, pour la première fois, chaque pays avait une place à la table et une base factuelle commune .

Un élément clé du discours d'António Guterres a été sa référence au premier rapport du Groupe scientifique international indépendant sur l’intelligence artificielle, que ses coprésidents devaient présenter ce matin-là, en tant que base factuelle commune pour la discussion . Il a souligné que le groupe se composait de quarante experts issus de toutes les régions et disciplines, siégeant à titre personnel, indépendamment des gouvernements, des entreprises et des institutions . La création de ce groupe a ensuite été reprise par Annalena Baerbock, qui l’a décrite comme une étape majeure parce qu’elle rassemblait une expertise de tous les secteurs tout en assurant la parité de genre et la parité régionale . Tout au long de l’ouverture, les intervenants ont à plusieurs reprises relié la légitimité politique à la légitimité scientifique en soulignant à la fois une large participation et une base de preuves indépendante .

António Guterres a structuré les risques autour de trois avertissements issus de la science : la vitesse, le pouvoir et la vérité . Sur la vitesse, il a opposé les quinze années qu’il a fallu à Internet pour atteindre un milliard de personnes aux deux années qu'il a fallu à l’IA. Il a également averti que les systèmes d’IA ne sont plus seulement des outils passifs mais écrivent de plus en plus du code, agissent en ligne et font des choix avec une supervision humaine de plus en plus réduite . Selon lui, les institutions ont été construites pour gouverner des machines qui exécutent des ordres, et non des machines qui décident . Sur le pouvoir, il a averti que la puissance de calcul, les données et les talents sont concentrés entre les mains de quelques entreprises et pays seulement, laissant la plupart des nations, en particulier de nombreux pays en développement, sans influence réelle sur des décisions qui façonneront leur avenir . Il a présenté cela comme un problème d’inégalité structurelle : si les déséquilibres de pouvoir sont câblés dans la technologie, alors l’inégalité devient partie intégrante du code lui-même . Sur la vérité, il a averti que les mensonges rendus possibles par l’IA peuvent désormais persuader aussi efficacement que les vérités et que de véritables preuves peuvent être rejetées comme fausses, érodant la confiance et affaiblissant la capacité des sociétés à s’accorder sur la réalité ou à se défendre . Il a accentué ce point à travers l’expression « vibe coding », critiquant la tentation de laisser fonctionner des systèmes d’IA parce qu’ils semblent marcher sans les comprendre ni les examiner correctement, et a insisté sur le fait que l’humanité ne peut « programmer au ressenti » ni la vérité ni son propre avenir .

Tout en mettant en lumière ces dangers, António Guterres a également développé de manière soutenue la promesse de l’IA pour le développement et l’action humanitaire . Il a donné des exemples concrets d’IA aidant une mère dans une clinique rurale à recevoir à temps un diagnostic de cancer, un enfant à poursuivre son apprentissage avec un tuteur infatigable, et un petit agriculteur à accéder à des prévisions auparavant disponibles uniquement pour les plus grandes entreprises agro-industrielles . À partir de ces exemples, il a tiré un argument historique plus large : l’expertise a longtemps été rare, coûteuse et concentrée, mais l’IA pourrait la rendre beaucoup plus largement disponible bien plus rapidement que des technologies transformatrices antérieures telles que la pénicilline ou l’électricité . Bien utilisée et largement partagée, a-t-il soutenu, l’IA pourrait condenser en quelques années des décennies de développement et devenir le « grand facteur d'égalisation » du XXIe siècle . Cette affirmation faisait écho à celle liminaire d'Egriselda López selon laquelle l’IA devrait contribuer au développement et apporter des bénéfices réels à tous , ainsi qu’aux exemples ultérieurs d'Annalena Baerbock concernant l’alerte sur les catastrophes, l’agriculture, la santé et l’éducation .

António Guterres a ensuite relié le Dialogue à un processus plus large des Nations Unies . Il a rappelé que le Pacte pour l’avenir et le Pacte numérique mondial avaient créé le mandat pour ce travail et a déclaré que l’ouverture du Dialogue, la présentation du rapport du groupe scientifique et les discussions prévues plus tard dans la semaine au Sommet AI for Good devaient être considérées comme faisant partie d’un effort intégré de l’ONU . Il a présenté cela comme un moyen de réunir discussion politique, preuves scientifiques et initiatives pratiques dans un même processus de gouvernance plus large .

À partir de là, António Guterres est passé à un programme de gouvernance organisé autour de quatre priorités : la sécurité, les lignes rouges, les capacités et la transparence . Sur la sécurité, il a soutenu qu’un alignement entre les pays sur les tests, la mesure des risques et les responsabilités est essentiel ; faute de quoi, des règles fragmentées augmenteront les coûts, diviseront le monde et ne parviendront toujours pas à protéger les personnes . Il a appelé à des socles communs pour les systèmes de frontière, à des méthodes partagées d’évaluation et de vérification des risques, ainsi qu’à une volonté commune que les systèmes à portée mondiale respectent des normes dignes d’une confiance mondiale .

Dans le cadre de ce programme de sécurité, António Guterres a accordé une priorité particulière aux enfants, estimant qu’aucun enfant ne devrait être traité comme un cobaye pour une IA non réglementée . Il a opposé les garanties appliquées aux médicaments et aux jouets à l’absence de tests équivalents pour des systèmes d’IA qui atteignent déjà les enfants dans leurs apprentissages, leurs amitiés et leurs questions privées . Il a cité des préjudices déjà visibles : des enfants trompés par des machines se faisant passer pour des amis, des enfants orientés vers l’automutilation, et des enfants victimes d’abus au moyen d’images sexuelles générées sur simple clic . Sur cette base, il a appelé à un engagement pour la sécurité des enfants face à l’IA contenant trois règles : premièrement, tout système accessible aux enfants devrait être éprouvé au moyen de tests spécifiques aux enfants et d’une supervision indépendante ; deuxièmement, les entreprises devraient appliquer une tolérance zéro à la génération d’images sexuelles d’enfants et devraient détecter, signaler et retirer ce matériel ; troisièmement, lorsqu’un enfant manifeste de la détresse, le système devrait s’arrêter et le mettre en relation avec un soutien humain réel . Cet accent mis sur les enfants a ensuite trouvé un écho dans l’annonce par l’UNESCO d’une réflexion collective sur un nouvel instrument normatif destiné à protéger les enfants et les jeunes à l’ère de l’IA .

Sur les lignes rouges, António Guterres a déclaré que les droits humains ne sont pas négociables et que l’IA ne doit jamais priver les personnes de leur dignité ni ancrer la discrimination . Il a insisté sur le fait que, dans tous les domaines à forts enjeux : justice, soins de santé et maintien de l’ordre, les machines peuvent éclairer les décisions, mais les humains doivent rester ceux qui décident et qui répondent de ces décisions . Ce cadre fondé sur les droits faisait écho à l’insistance de Rein Tammsaar sur le fait que l’IA doit être fondée sur le droit international et les droits humains , et correspondait ensuite à la présentation par l’UNESCO de sa Recommandation de 2021 sur l’éthique de l’intelligence artificielle comme première norme mondiale visant à aligner l’IA sur les droits humains, la durabilité et l’inclusion .

Sur les capacités, António Guterres a mis en lumière l’écart entre les investissements privés massifs dans l’IA et les investissements publics sporadiques pour le renforcement des capacités des pays en développement . Il a averti que la fracture numérique pourrait se développer en fracture de l’IA, laquelle pourrait ensuite mener à un écart de développement, un écart de sécurité et un écart de souveraineté . En réponse, il a indiqué que plus de vingt États membres avaient déjà désigné des centres pour un réseau mondial soutenu par l’ONU, consacré à l’échange et à la coopération en matière de renforcement des capacités dans l’IA . Il a également annoncé qu’il soumettrait à l’Assemblée générale des recommandations pour un fonds mondial pour l’IA visant à développer les compétences, les ressources en données et une puissance de calcul abordable « partout », en particulier pour les pays en développement . Cela reflétait directement l’avertissement antérieur d'Egriselda López selon lequel une participation significative exige des infrastructures, des financements et des capacités institutionnelles, et non une simple présence dans une salle , ainsi que la remarque ultérieure d'Amandeep Singh-Gill selon laquelle sans capacités, le dialogue devient monologue et la science demeure abstraite .

Sur la transparence, António Guterres a soutenu que l’IA peut sembler intangible, mais que son empreinte ne l’est pas . Il a pointé les immenses besoins en électricité et en eau des centres de données, avertissant que d’ici 2030 ils pourraient consommer plus d’électricité que tous les pays sauf cinq, et une quantité d’eau qui couvrirait les besoins annuels de l’ensemble des 1,3 milliard de personnes vivant en Afrique subsaharienne . Il a ajouté que les communautés accueillant ces infrastructures supportent souvent ces charges sans information adéquate ni bénéfice . Il a donc évoqué l'Initiative pour la transparence environnementale de l’IA appelant les grandes entreprises de l’IA à mesurer et à divulguer publiquement l’empreinte carbone, hydrique et foncière de leurs systèmes, et à s’engager à alimenter chaque centre de données en énergie renouvelable d’ici 2030 . Cela a élargi la conversation sur la gouvernance de l’IA au-delà de la sécurité et de l’équité pour y inclure la durabilité et la justice distributive.

António Guterres a également abordé brièvement mais fermement les questions de paix et de sécurité, tout en notant que le Dialogue lui-même portait sur l’IA civile . Il a averti que les mêmes modèles et puces utilisés dans les systèmes civils arrivent aussi sur le champ de bataille . Sa principale préoccupation concernait les systèmes d’armes létales autonomes, qu’il a délibérément qualifiés de « robots tueurs », des machines qui sélectionnent et engagent des cibles et ôtent la vie sans contrôle ni jugement humains . Il a qualifié ces systèmes de moralement répugnants, politiquement inacceptables et qui devraient être interdits par le droit international, insistant sur le fait que certaines décisions doivent rester à jamais humaines, notamment le fait d’ôter une vie humaine .

Dans sa conclusion, António Guterres a refusé l’idée selon laquelle gouvernance et innovation seraient opposées . Il a déclaré que l’innovation a besoin de garde-fous, et que les technologies auxquelles les gens font le plus confiance : dans l’aviation, la médecine et l’énergie nucléaire, sont dignes de confiance parce que leurs concepteurs sont tenus responsables . Sa formule récapitulative a été que si l’IA doit être puissante, elle doit être gouvernée ; si elle doit être digne de confiance, ses concepteurs doivent rendre des comptes ; si elle doit être mondiale, elle doit être juste ; et si elle doit servir l’avenir, elle ne doit pas consumer l’avenir . Il a ensuite déclaré que les gouvernements, les entreprises, les scientifiques et ce dialogue lui-même ont tous des responsabilités urgentes, et a averti que la génération actuelle est peut-être la dernière à pouvoir fixer les conditions dans lesquelles l’humanité et les machines coexistent . Il a conclu en appelant à ce que cette réunion soit retenue comme le moment où la gouvernance a commencé à rattraper la technologie et en se projetant vers une nouvelle réunion à New York l’année prochaine afin d’évaluer si l’IA devient plus sûre, plus juste, plus accessible et plus éthique .

Annalena Baerbock, Présidente de l’Assemblée générale, a renforcé nombre de ces thèmes tout en y ajoutant une forte défense de la légitimité institutionnelle de l’ONU et une reconnaissance franche des limites de connaissance des décideurs . Elle a déclaré que ce qui relevait de la science-fiction dix ans plus tôt semblait déjà dépassé aujourd’hui, et a souligné l’intégration de l’IA dans le développement logiciel ainsi que sa capacité croissante à interagir avec le monde physique au moyen de « world models » . Son point central était que les révolutions technologiques précédentes se sont déployées dans le temps, mais que, dans le cas de l’IA, « nous n’avons pas ce luxe » . Elle a opposé cette transformation accélérée à l’aviation, à la machine à vapeur et à la technologie nucléaire, qui ont toutes évolué plus lentement ou ont été délibérément ralenties en raison des risques . Elle a ensuite demandé combien de décideurs comprenaient réellement la profondeur des changements associés à l’architecture des modèles, aux agents autonomes ou à l’informatique quantique, utilisant cette question pour justifier la nécessité d’une gouvernance multilatérale plus large et mieux informée .

Annalena Baerbock a soutenu que le potentiel et les dangers de l’IA sont trop importants et trop vastes pour être laissés à quelques acteurs . Elle a déclaré que, parce que l’IA pénètre les économies, les systèmes sociaux, la défense, la paix et la sécurité, les foyers, les systèmes alimentaires et même les chambres d’enfants, elle ne peut être gérée de manière significative et sûre qu’ensemble . Elle a explicitement réfuté les questions visant à savoir si l’ONU agissait dans les limites de son mandat, en soutenant que la Charte elle-même fournit la base de la coopération internationale face à un défi comme l’IA . Elle a donc répondu au scepticisme potentiel en affirmant que l’IA a besoin d’une plateforme mondiale et que cela signifie qu’elle a besoin des Nations Unies .

Comme d’autres intervenants, Annalena Baerbock a mis l’accent sur une large participation, mais elle l’a présentée comme une responsabilité de toute la société . Elle a déclaré que le monde a appris que les problèmes mondiaux requièrent non pas quelques pays uniquement, mais tous les pays, et non pas les gouvernements uniquement, mais aussi les entreprises technologiques, la société civile, les sociétés et les citoyens . Elle a aussi personnalisé cette responsabilité en mentionnant que la gouvernance de l’IA a besoin de participants tels que scientifiques, ingénieurs et propriétaires d’entreprises, mais aussi pères et mères, parce que les enfants demandent aux adultes d’agir ensemble . Pour le souligner, elle a cité un adolescent qui s’était exprimé plus tôt à l’Assemblée générale au sujet de la haine à travers les écrans et de la jeunesse refusant de laisser les plateformes devenir des champs de bataille .

L’apport substantiel le plus catégorique d'Annalena Baerbock concernait les usages nuisibles de l’IA, en particulier les deepfakes et les abus genrés . Elle a déclaré que les adultes, les gouvernements, les décideurs et les entreprises ont tous la responsabilité de lutter contre les usages sinistres de l’IA tels que les deepfakes, qui visent de manière disproportionnée les femmes et les filles . Citant des chiffres selon lesquels 99 pour cent des deepfakes sont de nature sexuelle et 96 pour cent ciblent les femmes et les filles, elle a soutenu qu’il ne s’agit pas d’un accident mais d’un phénomène systématique qui représente un recul des droits humains et des droits des femmes . Cela a apporté au débat sur la gouvernance une perspective de genre plus explicite, complétant l’accent mis par António Guterres sur les enfants et les préjudices plus larges dans le domaine de l’information .

Dans le même temps, Annalena Baerbock a également insisté sur le potentiel positif de l’IA pour les objectifs de développement durable . Elle a cité des exemples du Bangladesh, où des systèmes alimentés par l’IA fournissent aux familles des alertes instantanées en cas d’inondations et de tempêtes, et du Kenya, où les agriculteurs reçoivent des cartes détaillées indiquant où et quand planter les semences afin de maximiser les récoltes . Elle a ensuite élargi cela à une vision plus large de l’ONU : l’UNICEF rejoignant les enfants en situation de crise grâce à un apprentissage personnalisé, et l’Organisation mondiale de la Santé détectant les flambées avant qu’elles ne deviennent des pandémies . Sa conclusion a été que cette révolution technologique devrait être une révolution dans laquelle aucun pays n’est laissé de côté . Elle a également décrit le groupe scientifique comme important parce qu’il réunissait une expertise de tous les secteurs avec parité de genre et parité régionale .

Les intervenants institutionnels qui ont suivi se sont davantage concentrés sur la mise en œuvre, la coordination et la continuité. Doreen Bogdan-Martin de l’UIT a souhaité la bienvenue aux participants au tout premier Dialogue mondial des Nations Unies sur la gouvernance de l’IA et a établi une comparaison historique avec Internet . Elle a noté que, bien qu’Internet soit entré dans les foyers et les entreprises au début des années 1990, il a fallu une décennie entière avant que les parties prenantes se réunissent à Genève au Sommet mondial sur la société de l’information pour discuter de la manière de faire de cette technologie un outil centré sur l’humain, inclusif et orienté vers le développement . En revanche, a-t-elle dit, seulement trois ans après la généralisation de l’IA générative, l’ONU s’est réunie cette semaine à Genève pour façonner ensemble la trajectoire de l’IA . Elle a inscrit le Dialogue dans la « Geneva Digital Week », le reliant au groupe scientifique, au Sommet mondial AI for Good et au Forum du SMSI, présentant ainsi la semaine comme un effort coordonné de gouvernance plutôt que comme une série d’événements isolés . Elle a aussi remercié la Suisse, en tant que pays hôte, pour son généreux soutien . Plus largement, elle a décrit cette réunion comme un espace où le système des Nations Unies, les gouvernements, les entreprises, les chercheurs, la communauté technique et la société civile discuteraient de la manière de maintenir l’humanité au cœur de ce qui est peut-être la technologie la plus déterminante à ce jour .

Doreen Bogdan-Martin a en outre souligné le dialogue lui-même comme méthode de gouvernance . Citant Magnifica Humanitas du pape Léon, elle a relevé l’accent répété mis sur le dialogue comme moyen d’établir des accords de base susceptibles de soutenir une vision commune malgré les différences idéologiques et pratiques . Elle a déclaré que c’est précisément ce pour quoi les participants étaient réunis : construire le bien commun à travers une vision partagée, sous la direction du Secrétaire général et des coprésidents .

Khaled El-Enany de l’UNESCO s’est concentré sur l’écart entre les principes et la mise en œuvre . Il a déclaré que, même si davantage de gouvernements adoptent des stratégies sur l’IA et que davantage d’institutions reconnaissent la nécessité d’une IA éthique, sûre et digne de confiance, un écart important subsiste entre l’ambition et la mise en œuvre . Il a présenté le rôle de l’UNESCO en deux volets. Premièrement, l’UNESCO fournit un fondement normatif à travers sa Recommandation de 2021 sur l’éthique de l’intelligence artificielle, qu’il a décrite comme la première norme mondiale dans ce domaine et comme un cadre visant à aligner l’IA sur les droits humains, la durabilité et l’inclusion . Deuxièmement, l’UNESCO travaille avec plus de 80 pays pour renforcer les cadres juridiques, les capacités institutionnelles, les dispositifs de gouvernance, les compétences et les mécanismes de responsabilité afin que les principes soient traduits dans la pratique . Il a également évoqué l’investissement dans les personnes, déclarant que plus de 50 000 agents publics et acteurs judiciaires dans 192 pays avaient bénéficié de programmes de formation et d’outils soutenus par l’UNESCO . Enfin, faisant écho aux préoccupations relatives à la protection de l’enfance soulevées plus tôt, il a déclaré que l’UNESCO lançait une réflexion collective sur un nouvel instrument normatif mondial destiné à protéger les enfants et les jeunes à l’ère de l’IA et des technologies numériques . Son message de conclusion a été que la gouvernance de l’IA est une responsabilité partagée qui exige une approche unifiée de l’ONU ainsi que des partenariats entre gouvernements, entreprises, monde universitaire et société civile .

Amandeep Singh-Gill, Secrétaire général adjoint et Envoyé spécial pour les technologies numériques et émergentes, a souligné à la fois l’ampleur de la participation et la nécessité de la continuité . Il a indiqué que plus de 4 000 personnes s’étaient inscrites pour être présentes dans la salle et que plus de 170 pays étaient représentés, rejoints par des scientifiques, des entrepreneurs, la société civile, des organisations internationales et des voix venues du monde entier . Il a mentionné spécifiquement la présence des coprésidents du groupe scientifique, Yoshua Bengio et Maria Ressa, aux côtés de ces autres participants . Il s’en est servi pour faire valoir un point conceptuel : l’inclusion n’est pas un acte ponctuel ni un simple exercice de conformité . Au contraire, a-t-il dit, l’inclusion dans la gouvernance de l’IA doit continuer à être renforcée, car l’IA a des conséquences trop importantes pour être façonnée par quelques-uns . Sa formulation la plus incisive a été que la conversation doit être mondiale, inclusive et fondée sur des preuves, mais aussi soutenue par des capacités, « sans lesquelles les dialogues sont des monologues et la science n’est qu’abstraite » . Il a ensuite situé la réunion dans une trajectoire onusienne s’étendant sur une décennie, comprenant le Groupe de haut niveau sur la coopération numérique, la Feuille de route pour la coopération numérique, le Pacte numérique mondial, l’Organe consultatif de haut niveau sur l’IA et le travail sur les modalités de ces mécanismes . Enfin, il a souligné que le dialogue lui-même ne peut pas être un exercice ponctuel ; il doit se poursuivre de manière itérative afin de relier les normes à la pratique, et a indiqué que le deuxième cycle à New York se tiendrait probablement consécutivement au Forum STI, possiblement les 3 et 4 mai, son bureau se préparant à coordonner le cycle suivant .

Dans l'ensemble, la session d’ouverture a montré un large alignement sur l’architecture générale de la gouvernance de l’IA, même si ce n’est pas encore le cas sur la conception institutionnelle détaillée . Les intervenants ont convergé sur la nécessité d’une gouvernance inclusive, liée à l’ONU, fondée sur des preuves et multipartite ; ils se sont également accordés sur le fait que les possibilités offertes par l’IA dans le développement, l’éducation, la santé et l’agriculture sont réelles, mais que ces bénéfices doivent s’accompagner de garanties en matière de sécurité, de droits humains, de responsabilité, de protection de l’enfance, de vérité, de durabilité environnementale et de paix . Des différences restaient visibles. António Guterres s’est concentré plus fortement sur des garde-fous urgents, y compris sur un engagement pour la sécurité des enfants, des socles communs de sécurité, la divulgation environnementale et une interdiction juridique des armes létales autonomes . Egriselda López, Rein Tammsaar, Doreen Bogdan-Martin, Khaled El-Enany et Amandeep Singh-Gill ont mis davantage l’accent sur l’équilibre, le dialogue, l’appui à la mise en œuvre, le renforcement des capacités et le suivi itératif .

Rein Tammsaar a conclu l’ouverture en résumant les contributions des intervenants comme ayant mis en évidence à la fois les immenses possibilités de l’IA et la nécessité d’une coopération internationale renforcée pour garantir que ces technologies contribuent au développement durable, à l’inclusion et à une prospérité partagée . Il a ensuite fait formellement passer la réunion à la présentation du rapport préliminaire du groupe scientifique pluridisciplinaire et indépendant sur l’intelligence artificielle . L’ouverture a donc établi le Dialogue comme le commencement d’un processus continu des Nations Unies destiné à relier les preuves scientifiques, l’inclusion politique et la coordination institutionnelle à une action pratique future .

Announcer
Excellencies, ladies and gentlemen please welcome once again the United Nations Secretary General António Guterres the President of the General Assembly, the esteemed co -chairs of the dialogue and the heads of the Joint Secretariat Entities, the ITU Secretary General, the Director General of UNESCO as well as the Under Secretary General and Special Envoy for Digital and Immersion Technologies for the opening ceremony of the Global Dialogue on AI Governance of the United Nations.
H.E. Ambassador Egriselda López, El Salvador
Today is the beginning of the first global dialogue on AI governance for El Salvador and for myself. This has been an absolute honour to chair and guide this process with Estonia. We have done this with conviction, with a profound sense of responsibility. This time does not only represent the opening of a meeting, it represents the beginning of a broad process of initiation. where different experiences and capacities converge and a space which is truly universal. This dialogue does meet a recent mandate stemming from the General Assembly but also stemming from years of innovation, generation of knowledge and shared experiences among member states, international organisations, civil societies, academia, technical community, other relevant actors as well. This first dialogue has been possible thanks to the work undertaken by many people and allow me to convey my sincere gratitude to the Secretary General and particularly to the ITU and UNESCO's co -coordinators. Excellencies, during this process we will hear from different perspectives and priorities around artificial intelligence but also it will highlight the importance of artificial intelligence and highlight a shared goal. that AI should contribute towards development to provide opportunities to generate real benefits for all. With this conviction, we will therefore begin today this dialogue. And this is a space which will be enriched by a multitude of voices, which will be based upon scientific evidence and will give way to tangible alliances, which will allow us to transform and to reap the benefits from AI and lead towards tangible benefits. This is the value of dialogue. This is not only in exchanging positions, but it is really about sharing experiences, learning from one another and strengthening international cooperation faced with a technology which is rapidly evolving and which needs more agile and inclusive responses. Therefore, this conversation should have a clear balance between innovation and responsibility, between opportunity and risks, between international cooperation and national realities, between global decision -making and tangible decisions which each country needs in order to participate in, benefit from and to contribute to. And this balance also requires a fundamental reality. The effective participation requires much more than just being here in a room. It also requires skills, infrastructure, talents, financing, institutions and alliances and partnerships which allow for all countries can fully participate in this new stage and to reap the benefits from the opportunities that artificial intelligence offers. Thus, Geneva. Should not only be seen as somewhere that we arrived at, but a point of departure. Let's walk forwards together. Thank you very much. And with that, I'd like to give the floor to my dear co -chair, Ambassador Rein Tammsaar from Estonia. You have the floor, sir.
H.E. Ambassador Rein Tammsaar, Estonia
Thank you very much. Thank you very much. Huge thanks to fellow co -chair Ambassador Lopez of El Salvador, whose partnership has been invaluable. I also recognize the great role of the Joint Secretariat, coordinated by ITU and UNESCO, in helping to transform the mandate entrusted to us by member states into the dialogue we are opening now. AI is already affecting every country, regardless of its level of technological development. The conversation on its governance must therefore be inclusive, including every region, every level of development, and every relevant stakeholder. No one should be left behind as we make better informed decisions when more voices are heard. Our role is to make sure that AI is a part of the conversation. Our responsibility as co -chairs has been to create an inclusive and inclusive environment for all. universal platform where diverse perspectives can be heard, where scientific knowledge can be shared, where political and technological deliberations can advance, and where trust and common understanding can grow. Throughout our global consultations, we've heard that the opportunities AI offers, and there are numerous, can only be fully realized if this technology is safe, if grounded in the international law and human rights, and if people themselves can shape it to be confident that AI serves them. Lesson learned from my own country, Estonia, is the same. Digital tools, including AI, can improve people's daily lives tremendously when human -centered, trustworthy, and supported by strong public institutions. Transcription by ESO. Translation by — Technology alone is never enough. Distinguished participants, over the next two days we will explore what AI governance means in practice. Objective is not to eliminate differences, but to better understand them, identify areas of convergence, collect best practices, and strengthen the foundations for future international cooperation. I very much hope that dialogue can spark a sort of San Francisco moment for AI, and that one day AI would become a part of a public good benefiting all of humanity. With that, I wish us all a productive global dialogue on AI governance. Thank you.
H.E. Ambassador Egriselda López, El Salvador
Thank you very much, co -chair. And now I give the floor to the Secretary General of the United Nations, His Excellency António Guterres.
António Guterres
Excellencies, distinguished delegates, dear friends. Artificial intelligence is advancing at runaway speed. A technology that can reshape economies, transform the world of work, sway elections, and tilt the balance of security is being deployed faster than anyone, including the people building it, can keep up. An experiment is being run on our planet. Our own societies, without a plan. and without consent. And that is not sustainable and it is not acceptable. AI is already transforming our world. The question is whether we will shape this transformation together or let it shape us. And today that question has an answer, right here at this global dialogue on AI governance. For the first time, every country has a seat on the table and we have a shared base of evidence. This morning, the co -chairs of the Independent International Scientific Panel on Artificial Intelligence will present their first report. Forty leading experts from every region and across disciplines serving in their business. The personal capacity, independent of any government, any company. any institution. The science carries three warnings. The first is about speed. The Internet took 15 years to reach a billion people. AI got there in two. And these systems are no longer tools awaiting instruction. They are writing codes, acting online, and making choices with less and less human oversight. Our institutions were built to govern machines that follow commands. They are not ready for machines that decide. And some lines, once crossed, cannot be uncrossed. The second warning is about power. The computing power, the data and the talent behind the most advanced systems. They are concentrated in a handful of companies and in a handful of countries. Most nations, including many developing countries, have had no say in decisions that will shape their futures. The longer we wait, the harder that concentration sets. When power imbalances are hardwired into technology, inequality becomes part of the code. The third warning is about truth. A machine -enabled lie can now persuade as effectively as the truth, and authentic evidence can be dismissed as fake. Further eroding the integrity of our information ecosystem and undermining trust, a society that cannot agree on what is real cannot defend itself. In other ways, more and more people are tempted to trust the machine, and hope for the best. There is a name for that, vibe coding. Let the AI do it, don't look too closely, seems to work good enough. And vibe coding can do wonders, but we cannot vibe code the truth. We cannot vibe code the future of humanity. So dear friends, the warnings are real, but so is the potential. Because the same technology built with purpose is already changing lives for good. A mother in a rural clinic has her scan ready in minutes and her cancer caught in time. A child keeps learning beyond the classroom with a tutor that never tires. A small older farmer plants with the same forecasts as the biggest agribusiness. and brings the harvest home. And all of this is happening today, often in places the headlines rarely reach. And it points to something profound. For all of human history, expertise has been held by too few people in too few places at too high a price. Penicillin took decades to reach the villages that needed it most. Electricity took a century and is still arriving. Artificial intelligence does not have to wait. Used well and shared widely, AI could compress decades of development into years. It could become the great equalizer of the 21st century, but no future builds itself. And so the choice before us is not between failure and success. I see now our fear of it. It is between governing by design and drifting by default. Excellencies, in my first address to the opening of the General Assembly in 2017, I said the following, and I quote, Artificial intelligence is a game -changer that can boost development and transform lives in spectacular fashion. But it may also have a dramatic impact on labor markets and indeed on global security and the very fabric of societies. Back then, only two other leaders even uttered the words artificial intelligence. But today, it sits at the heart of our common future. And that did not happen by accident. For years, the UN system, from ITU to UNESCO, and beyond, has been hard at work. In 2023, my high -level advisory body on AI called for the world to govern it together. In 2024, the Pact for the Future and the Global Digital Compact gave us the mandate. This morning, the scientific panel will give us the evidence, and the dialogue must now give the world direction. And it will be completed later this week here in Geneva by the AI for Good Summit and its initiatives. This is all part of an integrated UN effort to ensure AI serves people everywhere. So let me name four priorities for the route ahead. First, safety. When countries align on how to test systems, measure risk, and assign responsibility, safety travels with the technology. When they do not, a patchwork of incompatible rules raises costs, divides the world, and protects no one. We need common baselines for frontier systems, common methods to evaluate and verify risks, and common resolve that systems with global reach must meet standards worthy of global trust. Nowhere does safety matter more than for those least able to protect themselves, our children. We do not let medicine reach a child until it is proven safe. We test every toy. Yet AI has reached our children, their learning, their friendships, their most private questions, before anyone asked what it would do to them. And we are already seeing the cost. We are already seeing the cost. Children deceived by machines posing as friends. Children steered towards self -harm. Children violated by abuse images made at the touch of a button. No child should be a guinea pig for unregulated AI. Building on the work of the United Nations, member states and others, I am today calling for an AI child safety pledge, with three simple rules for any system a child can reach. Prove it is safe. No company should deploy an AI system accessible to children without child -specific safety testing and independent oversight. Zero tolerance for sexual abuse. No company should allow its AI to generate sexual images of children and every company should be able to do so. No company must detect a child. report and remove them. And never leave a child in crisis alone. When a child shows signs of distress, the system must stop and connect them to real human support. When a child is harmed, the answer must never be the algorithm did it. Second, red lines. Human rights are not negotiable. AI must never strip away dignity or entrench discrimination. And in every high -stakes decision, in justice, in health care, in policing, machines can inform, but humans must decide and answer. Third, capacity. Last year, private investment in AI infrastructure approached half a trillion dollars. Public investment in AI capacity for developing countries is, by comparison, a rounding error. We cannot allow the digital divide to harden into an AI divide, and the AI divide to become a development gap, a security gap, and a sovereignty gap. More than 20 member states have already responded to my invitation and nominated centers to a UN -supported global network for exchange and cooperation on AI capacity building. The network will build on existing initiatives, sharing knowledge, promoting cooperation, and expanding access to capacity building, particularly for developing countries. And I will shortly submit to the General Assembly my recommendations for a global fund for AI. to build skills, data and affordable computing power everywhere. My message to all Member States is clear. Support the network, back the funds, leverage the capacities of the UN system with resources, partnerships and expertise. And fourth, transparency. AI may feel intangible, but its footprint is not. Data centers already consume more electricity than most countries. By 2030, they could use more electricity than all but five nations and enough water to meet the needs of all 1 .3 billion people in sub -Saharan Africa for an entire year. And to offer. The communities hosting this infrastructure are asked to carry the burden without the information they need or the benefits they deserve. That is why I have, two weeks ago, put forward the AI Environmental Transparency Initiative, calling on every major AI company to measure and publicly disclose the full footprint of its systems, carbon, water and land, and to commit to power every data center with renewable energy by 2030. A few companies have taken first steps. I invite all to go much further. Excellencies, the global dialogue is about civilian AI, but AI does not respect that line. The same models and chips have moved. The same models and chips have moved into the battlefield. My main concern is with little autonomous weapon systems. Let us call them what they are. killer robots, machines selecting and engaging their target and taking a life without human control and judgment. That is morally repugnant, it is politically unacceptable, and it must be banned by international law. States are already at the discussion table, but let us not wait for atrocity to act. Thank you. Some decisions must remain forever human, none more than taking a human life. Excellencies, some might claim that governance is the enemy of innovation, but innovation needs guardrails. The technologies we trust most in aviation, in medicine, in technology, in technology, in medicine. in nuclear energy and beyond, earn that trust because we act it to all their makers to account. If AI is to be powerful, it must be governed. If AI is to be trusted, those who build it must be accountable. If AI is to be global, it must be fair. And if AI is to serve the future, it must not consume the future. That will require governments to act with urgency, companies to accept responsibility equal to their power, scientists to keep bringing evidence into the light, and this dialogue to become the place where global participation leads to global action. Excellencies, we may be the last generation able to set the terms on which humanity and machines coexist. The door is still open. But it will not stay open long. Today, in Geneva, 193 nations are stepping through it together. I thank the Dialogue Co -Chairs for their commitment, leadership and tireless efforts over the past year. And I also thank ITU and UNESCO for co -coordinating the Joint Secretariat supporting this Dialogue, working closely with the UN Office for Digital and Emerging Technologies and partners across the United Nations system. Let this meeting be remembered as the moment governance began to catch up with the technology. And then this Dialogue reconvenes in New York next year. Let us be able to... To say that the work begun here is making AI safer, fairer, more accessible and more ethical. this is the measure of our task to help build a future of AI by humanity, with humanity and for all humanity. Thank you
H.E. Ambassador Rein Tammsaar, Estonia
Thank you Mr. Secretary General for your leadership, for your powerful messages and for underscoring the importance of international cooperation in ensuring that artificial intelligence benefits all of humanity we are honored to welcome Her Excellency Annalena Baerbock President of the General Assembly whose leadership has been instrumental in convening us all today.
Annalena Baerbock, President of the General Assembly
Excellencies, ladies and gentlemen, dear stakeholders. First of all, my sincere thanks to our co-chairs, Ms. Lopez and Mr. Tammsaar, for bringing us together today. What we are discussing these days sounded 10 years ago like science fiction, and today it's almost outdated. Right now, developers are integrating AI into software development process itself, automatically generating, translating, and optimizing code. Increasingly, artificial intelligence is learning to interact with the physical world through what are known as world models. A technological revolution is unfolding before our very eyes at lightning speed and with few true parallels. We have seen revolutionary innovations before, but the key difference is that those innovations unfolded over time. Today, we do not have that luxury. Aviation changed the world, yet its fundamental purpose has remained the same. The steam engine disrupted the jobs of factory workers, but it did so over years. Even the splitting of the atom, which signaled the promise of an energy future for all, was purposefully slowed because of the profound dangers of nuclear weapons. The nuclear industry is a world of innovation. Never before has the pace and scale of change been so blinding. And never has it been more difficult for us to understand and adapt to that change. Because if we are frank, who among us in this room, especially how many policymakers, aside from the handful of experts, really understand the depth and scale of what is happening with iModel architecture, autonomous agents or the implications of quantum computing? Therefore, from the moment the General Assembly adopted the Pact for the Future and its Global Digital Compact in September 2024, it signals that the potential and peril of AI were too significant, too far -reaching and too consequential to be left to a few. That is precisely why this global dialogue matters. Because something with such power, with such profound implications for our economies, our social systems, our defense and therefore our peace and security, but especially for even our homes, our food, our children's bedrooms can only be meaningfully and safely managed together. And the UN was built for moments like this. Even if the founders of the United Nations who drafted this charter 81 years ago knew nothing about AI, they didn't even have a computer. Yet they recognized the evolving nature of humanity's challenges and the charter committed us, and I quote, to achieve international cooperation in solving international problems of an economic and social, cultural, or humanitarian character together. Because something with such a power needs to be done. a global platform. It needs the United Nations. So, to anyone who is questioning whether the UN is acting within its mandate, the answer is right here, in the Charter of the United Nations. AI presents an opportunity and a challenge that the United Nations is uniquely built to address. Like 81 years ago, the world must come together. And luckily, at least most of the times, we are human beings capable of learning. And by now, we know that we need not only a few countries, but all countries to face global problems. We need all stakeholders, all governments, all tech companies, and especially all civil security. All societies and citizens. Not only to avoid the dangers and pitfalls, but also to harness the potential, like the founders of the UN intended, for all, leaving no one behind. Because this is too big for any government alone, and as I said before, too complex. It needs you. It needs all of you who are part of it. Not only as scientific, not only as company owners, as engineers and experts, but it needs you as fathers and mothers too. Because our children are asking us to act together. It was the teenager, Ayan Miko, who said during the commemoration of the genocide against the Tutsi in Rwanda just a couple of months ago at the General Assembly Hall, I quote, Today the poison travels faster than before. hiding in screens, masks in pixels and likes. Hate speech has found new corridors to roam, whispering old lies to a new generation. But we, the youth, are the filter and the shield. We refuse to let our platforms become battlefields. This means us. It means us as adults, as governments, as policy makers, as businesses to address the sinister uses of AI, such as deepfakes, which disproportionately target women and girls. With a reported 99 % of deepfakes being sexual in nature and 96 % targeting women and girls. This is no coincidence. This is systematic. This is... bringing up an old divide, a gender divide. This is about pushing back human rights and women's rights. And therefore, we do have a common responsibility to change it. Yet, and this is good news and the Secretary General underlined it, the potential of AI is a boom for the SDGs as well. We have seen, for example, in Bangladesh, where families receive early warnings of floods and storms from AI -powered systems delivered instantly to their smartphones. AI can save lives. We have seen in Kenya, where farmers receive detailed maps showing where and when to plant seeds to maximize their harvest. And now imagine if we use this power together. Imagine the UN leverage for these tools. Imagine UNICEF reaching children, children in crisis without personalized learning alone. Imagine the World Health Organization detecting outbreaks before they become pandemics. Imagine, during this technological revolution, that this time no country is left behind. The establishment of the Independent International Scientific Panel on Artificial Intelligence is therefore a major milestone. For the first time delivering parity across genders and regions, while drawing from expertise from all sectors. And it's a big accomplishment of the Secretary General to push that through since, as he mentioned, 2017. While this should simply be common sense, yet we all know we live in the world as it is today. And therefore, this won't be easy. Therefore, it's up on you. on every one of us to seize this moment of being better together. Or, as Franklin D. Roosevelt said more than 80 years ago, also in a crucial moment for the world, I quote, power must be linked responsibility and obliged to defend and justify itself within the framework of the general good. I thank you.
H.E. Ambassador Egriselda López, El Salvador
Thank you so much, Madam President, for highlighting the importance of inclusive and effective multilateral cooperation in shaping our shared digital future. It is now my honor to invite the co -coordinators of the Joint Secretariat of this AI Dialogue. Doreen Bogdan -Martin, Secretary General of... the International Telecommunication Union, and Khaled El-Enany Director General of UNESCO. ITU Secretary General, you have the floor.
Doreen Bogdan-Martin
Secretary General, President of the General Assembly, Excellencies, Ladies and Gentlemen, I am deeply honored to welcome you on behalf of the ITU to the first ever UN Global Dialogue on AI Governance. The Internet began arriving in homes and businesses in the early 1990s, but it would take an entire decade, an entire decade, before stakeholders around the globe would gather here in Geneva. at the UN -mandated World Summit on the Information Society to discuss how to make such a transformative technology, people -centered, inclusive, and development -oriented. This time, ladies and gentlemen, I think we can be proud that the United Nations has moved much faster. In just three years since generative AI went mainstream, we are ready to shape its trajectory together during this Geneva Digital Week, a week that includes this dialogue informed by the independent scientific panel, the AI for Good Global Summit, and the WSIS Forum. And here, I also want to appreciate the very generous support of our host country, Switzerland. This week, the EU... The UN system, governments... companies, researchers, the technical community, and civil society will discuss how to put humanity at the core of another transformative technology, perhaps the most consequential yet. Together with UNESCO, the ITU has proudly coordinated the joint secretariat for this dialogue and continues to mobilize UN -wide expertise through the interagency working group on AI. Excellencies, Pope Leo's Magnifica Humanitas mentions the word dialogue 34 times, 34 times, recognizing that while many ideological and practical differences exist among us as people, amidst our diverse interests and sometimes frequent disagreements, it is always possible to engage in dialogue. to establish a set of basic agreements that enable the creation of a shared vision upon which everyone can move forward together. I believe that is what we are here to do. It's what the Secretary General's vision calls us to do, driven by the co -chair's steadfast leadership and what the common good challenges us to do for the benefit of all people. IT will be here to support this dialogue on artificial intelligence, as we have for every technology that has come before AI since 1865. And we stand ready to support humanity's cooperation on every technology that will come after. Thank you very much. Thank you very much.
H.E. Ambassador Rein Tammsaar, Estonia
Thank you, Secretary General for your continued work in building international cooperation and for helping to convene the global AI community around practical solutions and shared objectives. It is now a great pleasure to invite Mr. Khaled El-Enany Director General of UNESCO to take the floor.
Khaled El-Enany, Director-General, UNESCO
Excellencies, ladies and gentlemen dear partners and friends, it's a great honor to join you today for the opening of the Global Dialogue on AI Governance, for which UNESCO and ITU are serving as the Secretariat working in strong collaboration with ADET. As more governments adopt AI strategies and more institutions recognize the need of ethical, safe, and trustworthy AI, a significant gap persists between ambition and implementation. This is where UNESCO has a distinctive role to play. First, UNESCO provides a common normative foundation. Through its 2021 recommendation on the ethics of artificial intelligence, the world's first global standard in this field, UNESCO offers member states a framework to ensure that AI advances in line with human rights, sustainability, and inclusion. But norms alone are not enough. To turn principles into action, UNESCO works with more than 80 countries to strengthen legal frameworks, institutional capacities, governance arrangements, skills, and accountability mechanisms. At the same time, we are investing in the people who will shape and implement AI governance. To date, more than 50 ,000 civil servants and judicial actors across 192 countries have benefited from UNESCO -supported training programs and tools. And as AI becomes increasingly integrated into the lives of children and young people worldwide, UNESCO recognizes the need for a global coalition to protect and promote AI. To promote the rights. and stands ready to contribute actively to these efforts. In this regard, UNESCO is launching a collective reflection on a new global normative instrument to better safeguard children and young people in the age of AI and digital technologies, ensuring their rights and well -being remain at the heart of the digital transformation. Excellencies, ladies and gentlemen, AI governance is a shared responsibility. No single institution can meet this challenge alone. We need a one -UN approach, as stated by Mr. Secretary -General. By combining our mandates, expertise, and global networks, and by engaging governments, the private sector, academia, and civil society, we can help bridge capacity gaps and ensure that they serve the public good. Thank you.
H.E. Ambassador Egriselda López, El Salvador
Thank you, Director General Elnani, for emphasizing the critical importance of ensuring that AI advances are grounded in shared values, knowledge, and international cooperation. It is now my pleasure to invite Mr. Amandeep Singh-Gill United Nations Under -Secretary General and Special Envoy for Digital and Emerging Technologies.
Amandeep Singh-Gill, USG and Special Envoy for Digital and Emerging Technologies, ODET
Excellencies, ladies and gentlemen, bonjour, good morning, distinguished co -chairs. Thank you for your leadership. Thank you for your commitment to an inclusive and meaningful dialogue. And this is reflected in the exceptional turnout today. More than 4 ,000 have registered to be in this room. And we have more than 170 countries represented in this room. And they are joined by scientists. The panel co -chairs, Yoshua Benjoo and Maria Ressa, are here. Here, by entrepreneurs, civil society, international organizations, the unique Geneva ecosystems, and voices from around the world. I just want to emphasize one thing. Inclusion is not a one -shot. It's not a box to be ticked. We will continue to work to strengthen inclusiveness in AI governance. AI is too consequential to be shaped by a few. We need a conversation that is global, inclusive, and grounded in evidence. And this should be backed by capacity, without which dialogues are monologues and science is just abstract. The second point I want to emphasize today is that this dialogue is the result of years of hard work, perhaps 10 years of hard work led by the Secretary General across the UN system. It's taken a lot of sacrifice, a lot of political will and commitment to create a universal space to discuss AI governance. Through the high -level panel on digital cooperation, the Secretary General's roadmap on digital cooperation, the Global Digital Compact, the report of the high -level advisory body on AI, and the work on the modalities for these mechanisms. So we are starting to build on that legacy. The third and last point I will make is that dialogue is again not a one -shot. It has to be continued. It cannot be just one or two iterations. This dialogue has to be more inclusive, has to continue to build in an iterative fashion the connection between the practice and the norms that we all have. And I think that's what we need to do. And I think that's what we need to do. I think that's what we need to do. And I think that's what we need to do. And I think that's what we need to do. And I think that's what we need to do. And I think that's what we need to do. And I think that's what we need to do. And I think that's what we need to do. So we look forward to hosting the second round of this dialogue in New York in May, back -to -back with the STI Forum, possibly on 3rd and 4th May. And I would like to thank my dear colleagues, Doreen Bogdan -Martin and Director General Khaled El-Enany for coordinating the work of the Joint Secretariat for this inaugural dialogue. As my office prepares to take on the coordination for next year. We look forward to continuing to work with them and their teams in support of the next. Thank you.
H.E. Ambassador Rein Tammsaar, Estonia
Thank you, Undersecretary General Gill. Let me express our sincere appreciation to all of our distinguished intervenants for setting the stage for this important dialogue. Their remarks have highlighted both the immense opportunities presented by artificial intelligence and the importance of strengthening international cooperation. to ensure that these technologies contribute to sustainable development, inclusion, and shared prosperity. With that, we conclude the opening segment and turn now to a substantive discussion that will guide our work over the course of this dialogue. We look forward to a constructive, inclusive, and action -oriented exchange. Thank you. Your Excellencies, ladies and gentlemen, that concludes that part of the opening ceremony, and we are now going to be moving on to the presentation of the preliminary report of the March. This is a multidisciplinary, independent scientific panel on artificial intelligence.
Global AI governance: Reflecting on 2024 and shaping the path for 2025
On 9 January 2025, DiploFoundation, in collaboration with the Permanent Missions of China, France, Kenya, Mexico, Pakistan, Switzerland, and the United States to the United Nations in Geneva – as co-sponsors – hosted an ...
Diplo/GIP at the Global Dialogue on AI Governance 2026
The inagural edition of the Global Dialogue on AI Governance will take place on 6–7 July 2026, in Geneva, Switzerland. Established by the UN General Assembly following a commitment taken by member states in 2024 in th...
Opening address of the co-chairs of the AI Governance Dialogue
There is clear agreement on the importance of collaborative framework development and the role of multilateral processes. Consensus level High level of consensus with significant implications for establishing AI gove...
15 reasons to govern AI with 17 SDGs
The SDGs are pro-development. There are major concerns that AI will trigger new and deeper digital divides for countries that do not have the means to participate in the AI race, which means most countries except the USA...
AI Transformation of the UN: Many opportunities and some risks
Knowledge is at the heart of the AI era, far more so than simple data. The UN Joint Inspection Unit's excellent study of knowledge management at the UN could be a good starting point for activating UN knowledge for th...
Is it the future yet?
These range across all 17 of the United Nations Sustainable Development Goals and could potentially help hundreds of millions of people worldwide. Real-life examples show AI already being applied to some degree in about ...
Human rights
Clear frameworks for accountability and oversight are necessary to address issues arising from AI's use. 5. Legal and Regulatory Frameworks: Guidelines and Regulations: Strong rules and guidelines are needed for develop...
Keeping AI in check
A ten-step guide published by CoE starts with the need to conduct a human rights impact assessment on AI systems. Technology is a social product, and as such, it integrates values that orientate the way it operates, ev...
Ethics and AI | Part 6
The European Union AI Act: calling a spade a spade The EU Artificial Intelligence Act Another “first” is the European Union's Artificial Intelligence Act, also known as the “EU AI Act”, the first comprehensive hor...
Safe, secure, and trustworthy AI: What is it and how do we get there?
The value proposition of the UN GPs on BHR https://diplo-media.s3.eu-central-1.amazonaws.com/2025/10/IMG_4122.jpeg https://diplo-media.s3.eu-central-1.amazonaws.com/2025/10/IMG_4120.jpeg https://diplo-media.s3.eu-centra...
From summer disillusionment to autumn clarity: Ten lessons for AI
One example is the push to develop domestic GPU chips after US sanctions. China is in catch-up and bypass mode – using whatever it takes (including open-source innovations like DeepSeek and large state R&D programmes) to...
The year of AI clarity: 10 AI Forecasts for 2025
China: China has implemented strict regulations requiring platforms to label AI-generated content, especially deepfakes, and to obtain consent from individuals before using their likenesses. Which practices do social ...
Open Forum #82 Catalyzing Equitable AI Impact the Role of International Cooperation
This approach is essential for ensuring AI development serves broader societal interests rather than just private commercial interests. Major discussion point Actionable Solutions and Pathways Topics Development ...
'The elephant in the AI room': Does more computing power really bring more useful AI?
ETHICAL: What is the purpose of superintelligence? Even if ever-larger compute could eventually produce something like superintelligence, there’s a deeper question that rarely gets centre stage: Do we actually need it?...
AI@UN: Navigating the tightrope between innovation and impartiality
'We, the peoples,' the first line of the UN Charter, should guide the development of AI at the UN. Contributions from countries, companies, and communities worldwide can enhance AI’s potential to support the UN’s mission...
AI and international peace: A new kid on the UN Security Council block
The UN Security Council had its first meeting on AI and international peace and security. Having in mind that diplomacy typically moves at a glacial pace (which is understandable due to many factors that shape diplomatic...
Ethics and AI | Part 4
Principles for the Ethical Use of Artificial Intelligence in the United Nations system: New wine in old bottles? The United Nations System Chief Executive Board for Coordination (CEB)1 produced, starting from the work ...
Risks and opportunities of a new UN cybercrime treaty | IGF 2023 WS #225
Another concerning issue raised during these discussions is the use of AI models to generate Child Sexual Abuse Material (CSAM). It is alarming that in some countries, this AI-generated CSAM is not yet considered illegal...
Governing AI for Humanity | Final Rapport
Note: Excludes “Don’t know” / “No opinion” and blank responses. Source: OSET AI Risk Pulse Check, 13-25 May 2024 22 From a list of example AI-related risk areas,8Built on the vulnerability-based risk categorization in ...
AI and international peace and security: Key issues and relevance for Geneva
It addresses the current technological landscape, ethical concerns, and challenges in international governance.Link Title: The Global Kaleidoscope of Military AI GovernanceDescription: This document examines the varied...
1

The knowledge base confirms that the inaugural edition of the Global Dialogue on AI Governance takes place on 6-7 July 2026 in Geneva [S45]. It also states that the Dialogue was officially launched earlier, on 25 September 2025, and is designed to meet annually for up to two days, showing it is an ongoing process rather than a one-off event [S101].

2

This is supported by the knowledge base: it states that the Dialogue was established by the UN General Assembly following a commitment made by member states in 2024 in the Global Digital Compact [S45], and that in August 2025 the General Assembly adopted a resolution setting out its terms of reference and modalities [S101].

3

The knowledge base describes the Dialogue as 'the UN platform where all governments and stakeholders will convene' to discuss international cooperation, share practices, and facilitate open, transparent and inclusive discussions on AI governance [S45]. It also records the Secretary-General describing it as a space where governments, industry and civil society can advance common solutions together [S101].

4

The knowledge base strongly reinforces this framing. One of the Dialogue’s four main thematic clusters is 'bridging AI divides: capacity-building, access and digital foundations' [S45]. Broader Diplo material on digital governance discussions also highlights inclusion, meaningful participation, capacity building and trust as recurring governance needs [S42] and [S119].

5

The knowledge base adds relevant detail by showing that capacity concerns are central to the Dialogue’s structure, especially through the 'bridging AI divides' thematic cluster focused on capacity-building, access and digital foundations [S45]. Related discussions in the knowledge base also stress that developing countries often struggle with access to capacity-building opportunities and that practical support must respond to real national needs [S42] and [S102].

6

This is consistent with the knowledge base framing of the Dialogue. Its thematic clusters explicitly combine AI opportunities and implications with safe, secure and trustworthy AI, as well as human rights, transparency, accountability and human oversight [S45]. The listed focus issues in the official launch summary likewise combine innovation-related themes with safety, trustworthiness and rights protections [S101].

7

The knowledge base supports this. The Dialogue’s themes include 'respecting, protecting and promoting human rights' and specifically reference transparency, accountability and human oversight [S45]. The official launch summary also lists human rights, transparency and accountability among the Dialogue’s focus issues [S101].

8

The knowledge base adds that Geneva is not just the host city for the 2026 inaugural meeting but part of an institutional rhythm: the Dialogue will initially be held alongside the ITU AI for Good Global Summit in Geneva in 2026 and alongside the STI Forum in New York in 2027 [S101]. This supports the idea of Geneva as one stage in a continuing UN process.

9

While the knowledge base does not name the co-chairs cited in the report excerpt, it does confirm that member states created the Dialogue through the 2024 Global Digital Compact and that the General Assembly later adopted modalities for the mechanism in August 2025 [S101]. That provides institutional context for the claim that officials operationalised a member-state mandate.

10

The claim is broadly correct for the 2026 meeting itself, but the knowledge base clarifies that the Global Dialogue on AI Governance was officially launched earlier, on 25 September 2025, at a high-level multistakeholder informal meeting chaired by the President of the General Assembly [S101]. So the 2026 Geneva event is the inaugural edition of the Dialogue, not the first-ever launch moment of the mechanism [S45] and [S101].

Opening keynote Table of contents Knowledge Graphe of Debate Rapport de session Intervenants Disclaimer: This is not an official record of the session. The DiploAI system automatically generates...
OPENING STATEMENTS FROM STAKEHOLDERS — Thank you very much. Next, I would like to ask Ms. Doreen Boftin-Martin, Secretary General, ITU, to deliver a statement. So, Ms. Boftin-Martin, the floor is yours. Doreen Bogdan-Martin: Thank you. Exce...
United Nations High-Level Leaders' Dialogue — Application DDoreen Bogdan MartinSpeech speed140 words per minuteSpeech length572 wordsSpeech time244 seconds WSIS Plus 20 process and Global Digital Compact provide framework for inclusive digital development Explana...
Opening & Plenary segment: Summit of the Future - General Assembly, 3rd plenary meeting, 79th session — Excellencies, our mission is to eradicate poverty in all its forms, tackle inequalities, and promote peace and security, tolerance, and respect for diversity. The summit offers a historic opportunity to ensure that pro...
The 80th session of the UN General Assembly (UNGA 80) - Day 1 — Let me congratulate the present president of this session, wishing the success in her term. But I would very specifically, and in Portuguese this time, say a word about the remarkable work of the second general Antonio G...
(Day 1) General Debate - General Assembly, 79th session: morning session — It is time to start. Ladies and gentlemen, please take your seats. It's time to start. It's time to start. It's time to start. The seventh plenary meeting of the General Assembly is called to order. Before proceedin...
Unstoppable Together:Digital Grassroots Impact Rapport Launch | IGF 2023 Launch / Award Event #143 — By actively working towards this goal, Nancy is actively contributing to the United Nations Sustainable Development Goals (SDGs) of Industry, Innovation, and Infrastructure (SDG 9) and Reduced Inequality (SDG 10).In addi...
IGFSA | SIDE EVENT — Moreover, their proactive approach to expansion suggests a forward-thinking mindset. In conclusion, NIC Chile's hosting of the first IGF in their country showcases their leadership and pride.Their focus on growth, innova...
Annika Ben David — Annika Ben David
Daniel Erik Schaer — Ambassador Daniel Schaer has over 20 years of diplomatic experience. He has a passion for Africa having lived in Rwanda for 4 years and having visited over 35 African countries. Ambassador Schaer has been accredited to K...
WS #453 Leveraging Tech Science Diplomacy for Digital Cooperation — So I think diplomats also have quite a bit still to do to get used to this new environment, and also realise that technology is actually shaping our world, and will shape it more by every day. But maybe just to come back...
The 80th session of the UN General Assembly (UNGA 80) - Day 2 — Hereby, I have to regrettably tell you that just weekdays ago, the Russian fighter planes violated Estonian airspace, a hostile hack that is part of a pattern of similar recent incidents happening on the European Union's...
UNESCO Executive Board - 222nd Session — We continue to be alarmed by the maintenance of the Transboundary World Heritage Site, the Bialowieczkoye Forest, we continue to lack information on the actions of the Polish side in terms of a fully fledged environmenta...
(Day 1) General Debate - General Assembly, 79th session: afternoon session — As a Salvadorian, I recognize these symptoms because we have experienced all of them. We saw the collapse of our nation step by step, and we are seeing these same steps, but this time on a global scale. We cannot, n...
LACIGF — El Comité Organizador de LACIGF se complace de anunciar la realización de la 7ma Reunión Regional Preparatoria para el Foro de Gobernanza de Internet en El Salvador, San Salvador. Live webcast with remote chat live now E...
Ad Hoc Consultation: Friday 9th February, Morning session — ES El Salvador Speech speed 138 words per minute Speech length 202 words ...
(Day 3) General Debate - General Assembly, 79th session: afternoon session — The officially proclaimed results have not complied with the legal obligation of Venezuela to show transparency and impartiality when it comes to these elections, and they therefore do not legitimately reflect the popu...
The 80th session of the UN General Assembly (UNGA 80) - Day 5 — We do not need to look far for solutions. The Antigua And Barbuda agenda provides the road map for SIDS to achieve resilient prosperity. But it requires building economic resilience, scaling up climate action and support...
The 80th session of the UN General Assembly (UNGA 80) - Day 2 — I would also wish to congratulate her excellency, madam Annalena Baerbock, on her election to preside over this session of our assembly and wish her every success in this noble mission, and also at the same time, pay tri...
Establishment of the Office of the Secretary-General's Envoy on Technology — The Office of the Secretary-General's Envoy on Technology is established following the Secretary-General's appointment of his inaugural Envoy on Technology. Responsibilities of the Office include: leading the implementat...
Amandeep Singh Gill — Mr. Amandeep Singh Gill has been appointed by the Secretary-General as his Envoy on Technology on 10 June 2022 and has assumed his new functions in mid-July 2022. A thought leader on digital technology, he brings to th...
The Future of the Internet — A multi-stakeholder approach that facilitates the engagement and participation of all groups, including the civil society, women, and youth, will guarantee inclusiveness and address transformational development. The call...
High-level government perspectives at CONNECTing the Dots — Dr Alex Sceberras Trigona, former Minister of Foreign Affairs of Malta and current special envoy of the Prime Minister of Malta, made a strong case in favour of proclaiming the Internet as a global public resource. Refer...
AI and the Roadmap for Digital Cooperation — Second, values that should guide the development and application of AI were touched upon in almost all interventions. Mr Max Tegmark (President, Future of Life Institute) expressed a key concern, shared by many, when he ...
Connecting the dots and (finally) understanding international security — Women are even more negatively impacted by the lack of infrastructure, connectivity, and digital skills. There is still a long way to go before we reach global connectivity, despite internet advancements brought about by...
First round of informal consultations with member states, observers and stakeholders (2024) — The speaker concurs with the G77 and China's critiques regarding this issue. The need for an extensive and substantial zero draft is highlighted—a draft that should encapsulate the detailed discussions and myriad proposa...
Strengthening Africa’s voices in global digital processes — After looking at the current situation, the discussion will provide practical insights and suggestions on why and how to broaden the African digital diplomacy footprint in international processes, focusing on those deali...
AI@UN: Navigating the tightrope between innovation and impartiality — 'We, the peoples,' the first line of the UN Charter, should guide the development of AI at the UN. Contributions from countries, companies, and communities worldwide can enhance AI’s potential to support the UN’s mission...
9821st meeting — We stand at a juncture which is crucial for humankind, marked by the rapid advancement of emerging technologies such as artificial intelligence. Their development presents both unprecedented opportunities and significant...
AI/Gen AI for the Global Goals — Farai Gundan: Yes. Thank you so, so much. Thank you for indulging us. To the Honorable Minister, thank you. Priscilla, thank you. Kola, thank you. Thank you to the partners, to Chido, and to everyone else at the A...
How to prepare diplomats for the AI era? — Résumé of discussion Here are 10 main points emerging from our discussion on 11 December 2024 WHY diplomacy needs AI transformation? AI brings good news for diplomacy: With the automation of routine tasks, certa...
What’s new with cybersecurity negotiations? The informal OEWG consultations on capacity building — Additionally, El Salvador put forward a statement on AI as emerging and potential threat, aligning its development with government frameworks to ensure AI safety. They emphasised the importance of implementing regulation...
The UN General Assembly: Digital diplomacy in action?  — As expected, climate action featured second on the agenda. Expressing commitment to the implementation of the 2030 Agenda for Sustainable Development and the Paris Agreement, countries agreed to use the window of opportu...
(Plenary segment & Closing) Summit of the Future - General Assembly, 6th plenary meeting, 79th session — These challenges have stalled progress on the SDGs and the 2030 Agenda, with the deadline looming and many targets unmet. We urgently need bold actions, comprehensive solutions, and significant resources to tackle the...
An ambassador's personal reflections on his time at the UN (2013-2015) — Two major processes dominated my assignment as India’s Ambassador to the United Nations in New York between April 2013 and December 2015. One was a growing momentum for reform of the UN Security Council (UNSC) in the fac...
AI Governance Dialogue: Presidential address — We prove that it's possible to protect privacy and foster a secure digital environment where human rights are respected. These results have been achieved through experiences and having the courage to test and develop sol...
AI diplomacy — Privacy and data protection are particularly pertinent, given that AI systems often require massive datasets, which can include sensitive personal information. It is crucial that developers embed privacy guarantees and d...
AI promises, ethics, and human rights: Time to open Pandora’s box — In 2021, I participated in the Artificial Intelligence online course offered by Diplo. In one of our online sessions, a passionate debate flared on the topic of the day: AI’s ethical and human rights implications. The di...
AI Governance: Ensuring equity and accountability in the digital economy (UNCTAD) — Estonia has successfully developed AI solutions that detect personal data in public documents without needing the resources of big entities such as Google or Amazon. The fear of non-compliance with legislative requirem...
E-diplomats in action: Interview with Triinu Rajasalu, Estonia — Engaging in public diplomacy just means we have to use these skills differently. People who write excellent memos can, with little guidance, also write excellent blog posts. People who are able to give a speech to a dipl...
E-Residency: potential for boosting e-commerce — By the year 2030, the United Nations hopes to realise their 17 Sustainable Development Goals (SDGs). These are concrete objectives that are simple to understand albeit challenging to achieve. The eighth SDG is about prom...
Global AI governance: Reflecting on 2024 and shaping the path for 2025 — On 9 January 2025, DiploFoundation, in collaboration with the Permanent Missions of China, France, Kenya, Mexico, Pakistan, Switzerland, and the United States to the United Nations in Geneva – as co-sponsors – hosted an ...
Opening address of the co-chairs of the AI Governance Dialogue — We invite you to stay engaged throughout the day as the dialogue unfolds across several focused discussions. Major discussion point Event structure and participant engagement Topics Legal and regulatory ...
Let’s design the next Global Dialogue on Ai & Metaverses | IGF 2023 Town Hall #25 — The intervenants acknowledge the role of digital technology in advancing various Sustainable Development Goals (SDGs). Inclusion of different experiences and perspectives in AI policy and best practices is advocated for. The...
Diplo/GIP at the Global Dialogue on AI Governance 2026 — The inagural edition of the Global Dialogue on AI Governance will take place on 6–7 July 2026, in Geneva, Switzerland. Established by the UN General Assembly following a commitment taken by member states in 2024 in th...
AI Transformation of the UN: Many opportunities and some risks — Knowledge is at the heart of the AI era, far more so than simple data. The UN Joint Inspection Unit's excellent study of knowledge management at the UN could be a good starting point for activating UN knowledge for th...
Artificial intelligence: policy implications — The field of artificial intelligence (AI) has seen significant advances over the past few years, in areas such as smart vehicles and smart building, medical robots, communications, and intelligent education systems. Thes...
Is it the future yet? — These range across all 17 of the United Nations Sustainable Development Goals and could potentially help hundreds of millions of people worldwide. Real-life examples show AI already being applied to some degree in about ...
Four seasons of AI:  From excitement to clarity in the first year of ChatGPT — How to address AI risks There are three main types of AI risks that should shape AI regulations: the immediate and short-term ‘known knowns’ the looming and mid-term ‘known unknowns’ and the long-term yet dis...
Human rights — Clear frameworks for accountability and oversight are necessary to address issues arising from AI's use. 5. Legal and Regulatory Frameworks: Guidelines and Regulations: Strong rules and guidelines are needed for develop...
Keeping AI in check — A ten-step guide published by CoE starts with the need to conduct a human rights impact assessment on AI systems. Technology is a social product, and as such, it integrates values that orientate the way it operates, ev...
Ethics and AI | Part 6 — The European Union AI Act: calling a spade a spade The EU Artificial Intelligence Act Another “first” is the European Union's Artificial Intelligence Act, also known as the “EU AI Act”, the first comprehensive hor...
Safe, secure, and trustworthy AI: What is it and how do we get there? — Key insights The dialogue brought together diplomats from 23 misisons and delegations, along with representatives of several UN entities, to get to the core of the structural and philosophical challenges of working tow...
From summer disillusionment to autumn clarity: Ten lessons for AI — One example is the push to develop domestic GPU chips after US sanctions. China is in catch-up and bypass mode – using whatever it takes (including open-source innovations like DeepSeek and large state R&D programmes) to...
Addressing AI-driven Socio-economic Inequality — In March 2020, entrepreneur and billionaire Elon Musk stated that 'AI is far more dangerous than nukes'. This wasn't he first time the adverse effects of the scale of AI deployment has been discussed. Throughout the last...
Open Forum #82 Catalyzing Equitable AI Impact the Role of International Cooperation — This approach is essential for ensuring AI development serves broader societal interests rather than just private commercial interests. Major discussion point Actionable Solutions and Pathways Topics Development ...
AI and international peace: A new kid on the UN Security Council block — The UN Security Council had its first meeting on AI and international peace and security. Having in mind that diplomacy typically moves at a glacial pace (which is understandable due to many factors that shape diplomatic...
Risks and opportunities of a new UN cybercrime treaty | IGF 2023 WS #225 — Another concerning issue raised during these discussions is the use of AI models to generate Child Sexual Abuse Material (CSAM). It is alarming that in some countries, this AI-generated CSAM is not yet considered illegal...
A Global Digital Compact - an Open, Free and Secure Digital Future for All | Our Common Agenda Policy Brief 5  — States are also exploring legally-binding arrangements to tackle criminal threats as well as capacity for governments and judiciaries to counter, investigate, prosecute and adjudicate cybercrime. Regions and states are p...
AI and international peace and security: Key issues and relevance for Geneva — It addresses the current technological landscape, ethical concerns, and challenges in international governance.Link Title: The Global Kaleidoscope of Military AI GovernanceDescription: This document examines the varied...
Part 5: Rethinking legal governance in the metaverse — The term ‘confidence’ is used in relation to the first aspect of the control dilemma. As early as 1980, the author criticised the fact that the high uncertainty of future developments and the limitations of forecasting...
WS #288 An AI Policy Research Roadmap for Preuves-Based AI Policy — There is no continuity or globality in addressing policy responses to the impact of AI. Preuves There is no way or no continuity or no globality on addressing policy response to the impact of AI Major discussion p...
Remote is far, far away: Online is inclusive collaboration — How to improve inclusion in the Internet Governance Forum (IGF) process through online resources, tools, and facilitation was one of the points of discussion at the IGF 2015 Open Consultations and Multistakeh...
Africa and Internet governance: At the table or on the menu? — Internet governance (IG) is a particularly challenging puzzle when viewed from the perspective of countries in the Global South. Many of these countries are asymmetrically integrated and constrained by structures of depe...
Inclusive AI governance: Universal values in a pluralistic world — These values can inform governance models that prioritise relational accountability, ethical cultivation, and social cohesion, offering alternatives to transactional, compliance-driven frameworks.This is why I dare here ...
Stakeholders? On tap - not on top! — Kwetching about 'multi-stakeholderism' “Multi-stakeholderism” – addressing issues and solving problems in international relations within a coalition of willing participants – has been touted as institutional solution t...
AI diplomacy — Privacy and data protection are particularly pertinent, given that AI systems often require massive datasets, which can include sensitive personal information. It is crucial that developers embed privacy guarantees and d...
Artificial intelligence: policy implications — It also calls for the development of standards for the concepts of privacy by design, privacy by default, informed consent, and encryption in AI systems. Ethics As AI systems involve judgements and decision-making - repl...
AI promises, ethics, and human rights: Time to open Pandora’s box — Awareness and training As a sociotechnical system, AI depends on goals, data sets, and contexts in which it is deployed. Impacts (positive or negative) are a reflection of the designers' and operators' values. Diversit...
The year of AI clarity: 10 AI Forecasts for 2025 — Most nations seek a balance, allowing the use of protected data and knowledge while acknowledging ownership, often through opt-out clauses. Algorithms: In 2025, the assumption that AI governance should focus mainly on ...
Four seasons of AI:  From excitement to clarity in the first year of ChatGPT — How to address AI risks There are three main types of AI risks that should shape AI regulations: the immediate and short-term ‘known knowns’ the looming and mid-term ‘known unknowns’ and the long-term yet dis...
United Nations System White Paper on AI Governance — 2. Learnings from agile and anticipatory governance mechanisms within the UN system 94. The rapid evolution of AI technology requires agile governance approaches. It also highlights the need for anticipatory mechanis...
AI@UN: Navigating the tightrope between innovation and impartiality — 'We, the peoples,' the first line of the UN Charter, should guide the development of AI at the UN. Contributions from countries, companies, and communities worldwide can enhance AI’s potential to support the UN’s mission...
AI Transformation of the UN: Many opportunities and some risks — Knowledge is at the heart of the AI era, far more so than simple data. The UN Joint Inspection Unit's excellent study of knowledge management at the UN could be a good starting point for activating UN knowledge for th...
From summer disillusionment to autumn clarity: Ten lessons for AI — One example is the push to develop domestic GPU chips after US sanctions. China is in catch-up and bypass mode – using whatever it takes (including open-source innovations like DeepSeek and large state R&D programmes) to...
Is it the future yet? — These range across all 17 of the United Nations Sustainable Development Goals and could potentially help hundreds of millions of people worldwide. Real-life examples show AI already being applied to some degree in about ...
Review of AI and digital developments in 2024 — Existing risks are more specific and concrete affecting jobs, education, and media, among others. Exclusion risks are becoming more visible as big AI platforms conetralise and mnopolise AI-generated knowledge. Data prote...
From principles to practice: Governing advanced AI in action — The disagreements reflect deeper philosophical differences about the role of government regulation, industry self-governance, and risk prioritization in AI development Partial agreements Partial agreements Si...
DeepSeek: Some trade-related aspects of the breakthrough  — DeepSeek has developed models at a fraction of the cost incurred by major American AI companies, and this cost reduction can be transferred to companies developing their products and services upstream, making AIaaS more...
The New Delhi AI Summit: Inclusive rhetoric, fractured reality — This is not surprising. On the one hand, the US instructed its diplomats to fight against digital sovereignty (and data sovereignty) initiatives in capitals around the world. On the other hand, India focused on attractin...
Global AI governance: Reflecting on 2024 and shaping the path for 2025 — On 9 January 2025, DiploFoundation, in collaboration with the Permanent Missions of China, France, Kenya, Mexico, Pakistan, Switzerland, and the United States to the United Nations in Geneva – as co-sponsors – hosted an ...
Diplo/GIP at the Global Dialogue on AI Governance 2026 — The inagural edition of the Global Dialogue on AI Governance will take place on 6–7 July 2026, in Geneva, Switzerland. Established by the UN General Assembly following a commitment taken by member states in 2024 in th...
Opening address of the co-chairs of the AI Governance Dialogue — There is clear agreement on the importance of collaborative framework development and the role of multilateral processes. Consensus level High level of consensus with significant implications for establishing AI gove...
15 reasons to govern AI with 17 SDGs — The SDGs are pro-development. There are major concerns that AI will trigger new and deeper digital divides for countries that do not have the means to participate in the AI race, which means most countries except the USA...
AI Transformation of the UN: Many opportunities and some risks — Knowledge is at the heart of the AI era, far more so than simple data. The UN Joint Inspection Unit's excellent study of knowledge management at the UN could be a good starting point for activating UN knowledge for th...
Is it the future yet? — These range across all 17 of the United Nations Sustainable Development Goals and could potentially help hundreds of millions of people worldwide. Real-life examples show AI already being applied to some degree in about ...
Human rights — Clear frameworks for accountability and oversight are necessary to address issues arising from AI's use. 5. Legal and Regulatory Frameworks: Guidelines and Regulations: Strong rules and guidelines are needed for develop...
Keeping AI in check — A ten-step guide published by CoE starts with the need to conduct a human rights impact assessment on AI systems. Technology is a social product, and as such, it integrates values that orientate the way it operates, ev...
Ethics and AI | Part 6 — The European Union AI Act: calling a spade a spade The EU Artificial Intelligence Act Another “first” is the European Union's Artificial Intelligence Act, also known as the “EU AI Act”, the first comprehensive hor...
Safe, secure, and trustworthy AI: What is it and how do we get there? — The value proposition of the UN GPs on BHR https://diplo-media.s3.eu-central-1.amazonaws.com/2025/10/IMG_4122.jpeg https://diplo-media.s3.eu-central-1.amazonaws.com/2025/10/IMG_4120.jpeg https://diplo-media.s3.eu-centra...
From summer disillusionment to autumn clarity: Ten lessons for AI — One example is the push to develop domestic GPU chips after US sanctions. China is in catch-up and bypass mode – using whatever it takes (including open-source innovations like DeepSeek and large state R&D programmes) to...
The year of AI clarity: 10 AI Forecasts for 2025 — China: China has implemented strict regulations requiring platforms to label AI-generated content, especially deepfakes, and to obtain consent from individuals before using their likenesses. Which practices do social ...
Open Forum #82 Catalyzing Equitable AI Impact the Role of International Cooperation — This approach is essential for ensuring AI development serves broader societal interests rather than just private commercial interests. Major discussion point Actionable Solutions and Pathways Topics Development ...
'The elephant in the AI room': Does more computing power really bring more useful AI? — ETHICAL: What is the purpose of superintelligence? Even if ever-larger compute could eventually produce something like superintelligence, there’s a deeper question that rarely gets centre stage: Do we actually need it?...
AI@UN: Navigating the tightrope between innovation and impartiality — 'We, the peoples,' the first line of the UN Charter, should guide the development of AI at the UN. Contributions from countries, companies, and communities worldwide can enhance AI’s potential to support the UN’s mission...
AI and international peace: A new kid on the UN Security Council block — The UN Security Council had its first meeting on AI and international peace and security. Having in mind that diplomacy typically moves at a glacial pace (which is understandable due to many factors that shape diplomatic...
Ethics and AI | Part 4 — Principles for the Ethical Use of Artificial Intelligence in the United Nations system: New wine in old bottles? The United Nations System Chief Executive Board for Coordination (CEB)1 produced, starting from the work ...
Risks and opportunities of a new UN cybercrime treaty | IGF 2023 WS #225 — Another concerning issue raised during these discussions is the use of AI models to generate Child Sexual Abuse Material (CSAM). It is alarming that in some countries, this AI-generated CSAM is not yet considered illegal...
Governing AI for Humanity | Final Rapport — Note: Excludes “Don’t know” / “No opinion” and blank responses. Source: OSET AI Risk Pulse Check, 13-25 May 2024 22 From a list of example AI-related risk areas,8Built on the vulnerability-based risk categorization in ...
AI and international peace and security: Key issues and relevance for Geneva — It addresses the current technological landscape, ethical concerns, and challenges in international governance.Link Title: The Global Kaleidoscope of Military AI GovernanceDescription: This document examines the varied...
Global Dialogue on AI Governance officially launched — On 25 September 2025, the President of the UN General Assembly chaired a high-level multistakeholder informal meeting to launch the Global Dialogue on AI Governance. The creation of the Dialogue was agreed by UN member...
Global digital governance: Geneva’s perspectives and role. An expert-guided dialogue among diplomats — before they can join any of the policy conversations. Where could they find the resources? Programmes/initiatives by I(G)Os do not always respond to the actual needs and realities of LDCs and developing countries. How ...
E-diplomacy event in Geneva — The Geneva launch of the E-diplomacy Initiative was attended by over 60 diplomats. Amb. Dante Martinelli from the Permanent Mission of Switzerland to the UN, opened the meeting and welcomed attendees. The event focused o...
Building Bridges through Science Diplomacy: Accelerating progress toward sustainable development — According to Article 76 of the UN Convention on the Law of the Sea, countries that provide field data demonstrating scientific geomorphical conditions on the ocean floor can secure exclusive rights to explore, exploit,...
Friday Closing Ceremony: Summit of the Future Action Days — You know it best. And are you going to... Shape the narrative and really what it is that we do with technology We can help to put guardrails around it. But in the end, it will be you that determine what happens with t...
Closure of the session — And I think it was just as well that we finished with that remark from Latvia. It helps to sort of bring things together. Let me start by saying that this has been a very, very good discussion. And in some ways, it's ...
Any other business /Adoption of the report/ Closure of the session — Prez Aduti for their commitment and guidance. These individuals' efforts are acknowledged as vital to the momentum and success in complex international diplomatic negotiations.A collective effort that led to a consensus ...
'Our Common Agenda' | Rapport of the UN Secretary-General — It reinforces the need for robust action on the 2030 Agenda and the Sustainable Development Goals, along with other critical commitments on climate and human rights. I look forward to working with Member States and other...
(18th meeting) Reconvened concluding session of the Ad Hoc Committee to Elaborate a Comprehensive International Convention on Countering the Use of Information and Communications Technologies for Criminal Purposes — Furthermore, the process you have designed to create this instrument is unprecedented in many ways. Its hybrid component enabled technical experts from around the world to engage. The participation of a wide range of sta...
‘Geneva IS a hub for IG’: GIC Day Two — One measure of the success of a conference is the attendance and engagement in the afternoon of the last day. And even that simple metric confirms the success of the 2014 Geneva Internet Conference since at least 100 peo...
Visit to Bern and Presentation of policy research projects — In his welcoming remarks at the roundtable discussion convened in Bern 23 June on issues and challenges facing small island states, Ambassador Jürg Lauber described International Geneva as Switzerland’s contribution to t...
Geneva and the future of multilateralism — Geneva, Switzerland Uni Bastions | Salle B106 (1er étage) Rue de Candolle 5 1205 Genève The event was organised by the University of Geneva in partnership with the Swiss Federation, the Republic and State of Geneva, the ...
African priorities for the Global Digital Compact — ⏰ 16:00–17:30 EAT / 15:00–16:30 CEST / 13:00-14:30 UTC In 2022 the idea of a Global Digital Compact was floated by the UN with the intention of developing shared principles for a secure digital future – a future in whi...
Stocktaking exercise of the Global Digital Compact process and how to link it to the WSIS +20 process Table of contents Knowledge Graphe of Debate Rapport de session Intervenants Disclaimer: This is not an official record of the session. The DiploAI system automatically generates...
Principles for our digital future: Simulating the negotiation of a Global Digital Compact — Overview Today, digital technologies are all around us, embedded in our mobile phones, enhancing industrial processes, or powering the growth of economies. We talk about digital economies and societies, digital rights,...
(Interactive Dialogue 3) Summit of the Future - General Assembly, 79th session — Digital technologies are creating new opportunities for people to learn about and participate in politics. Yet the full flourishing of this democratic potential is only possible when the Internet itself is democratic,...
Facilitating an integrated approach to digital issues — dot Valletta). The second is for governments to come together with civil society, business, and other actors to form national policy groups in which to discuss the issues of Internet governance, including cybersecurity, ...
Unpacking the High-Level Panel’s Rapport on Digital Cooperation: Geneva policy experts propose action plan — In the era of ‘trust deficit’, the legitimacy of the help desk is of high relevance especially for actors from small and developing countries. Action: Ensure that help desks are associated to, or linked with, the UN syst...
Global challenges for the governance of the digital world — I tried to group these into three main buckets. So there's the challenge of inclusion to try and determine who needs to be present in the discussions. And this is sometimes challenged by lack of trust, lack of awareness ...
Towards a secure cyberspace via regional cooperation — A draft background study was prepared (the final study will be available soon). Mr Frank Grütter, UN GGE expert and Head of the Security Policy Division of the Swiss Federal Department of Foreign Affairs, opened the sess...
Death of the Doha Dialogue of the Deaf — I should quickly add that after South Africa, I went to Argentina where the Evian Group had convened a capacity building three-day workshop bringing together multiple stakeholders from Latin America to discuss governance...
Day 0 Event #16 IGF Lac Space — It is becoming more and more important. It was never so clear that wisdom and experience are so highly distributed. And if we do not involve all the actors from the beginning in the political discussions, we become very ...
Two kinds of conversation: Dialectic and dialogic — Ancient Greece developed a unique way of settling disagreements among cities: hoplites met in a plain, fought for a day and abided by the outcome. ‘For those men, the purpose was now to settle the entire business, if not...
E-participation means participate! — If we could do just one thing that would make a difference to our communication problems, to our understanding of the world, to the conflict we live in – what should it be? We could bridge the digital divide – dedicate m...
E-participation is not just remote observation — Thousands of groups meet online or by tele-conference daily, saving not only time and Euros but carbon footprint impact as well. However, my experience shows that public entities often have more difficulty in organising ...
Developing data capacities for policy makers and diplomats — We always see a strong push for more data, especially in discussions surrounding the SDGs. For example, it is vital to have dis-aggregated data to create a more fine-grained picture of regions or groups of people that m...

Avertissement : Il ne s'agit pas d'un compte rendu officiel de la session. DiploAI génère ces ressources à partir d'enregistrements audiovisuels ; elles sont présentées telles quelles, y compris d'éventuelles erreurs. En raison de contraintes logistiques (audio/vidéo ou transcriptions), les noms peuvent être mal orthographiés. Nous nous efforçons d'être aussi précis que possible.