Chinese AI models are narrowing the gap with leading US systems, according to a new analysis by the Center for Strategic and International Studies.
CSIS said recent releases from Z.ai, Moonshot, DeepSeek and Alibaba-backed Qwen show that China’s rapid progress in AI was not limited to DeepSeek-R1, but reflects a broader pattern of fast technical catch-up.
The analysis points to Z.ai’s GLM-5.2 model, which performs close to the top US closed models in coding and agent-based tasks. It also highlights strong results from Moonshot’s Kimi, DeepSeek V4-Pro and Qwen3.7-Max across software engineering, reasoning and agent benchmarks.
CSIS argues that Chinese models are now only months, rather than years, behind US frontier systems in several practical areas.
The report identifies knowledge distillation, open-weight research communities and efficiency-driven engineering as key factors behind this progress. Chinese labs can learn quickly from stronger models, shared research practices and open-source ecosystems, while US chip export controls have pushed them towards more efficient training and inference strategies.
Cost is another important factor. CSIS said Chinese models are often cheaper to access than leading US closed systems because open-source releases can be hosted by many providers, increasing price competition and making them easier for developers and governments to adopt.
The analysis says US firms still retain major advantages in frontier capabilities, cloud platforms, enterprise products and user feedback loops. However, Chinese models are now capable, affordable and open enough to shape global AI competition.
CSIS argues that US policy should therefore focus not only on protecting technological advantage, but also on building global trust, lowering access costs and ensuring partners see the American AI stack as reliable.
Why does it matter?
The analysis shows that AI competition is not only about which country has the most powerful frontier model. Chinese open-weight models are spreading because they are increasingly capable, cheaper to run and easier to deploy through third-party hosts or local infrastructure. That could shape global adoption, especially for governments, startups and developers that cannot afford or do not want to depend entirely on US closed-model providers. For the US, the challenge is no longer only maintaining a technical lead, but also making its AI ecosystem trusted, affordable and reliable for international partners.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The multidisciplinary Independent International Scientific Panel on Artificial Intelligence presented its first annual report during the United Nations Global Dialogue on AI Governance in Geneva, offering an evidence-based assessment of AI’s opportunities, risks and societal impacts.
The session formed part of the inaugural Global Dialogue on AI Governance, held on 6-7 July. The dialogue was established in 2025 to support open, transparent and inclusive discussions on international AI governance, including AI’s role in sustainable development, digital divides, safety, human rights, transparency, accountability and human oversight.
Opening the presentation, Yoshua Bengio, professor of computer science at the University of Montreal, said the panel’s role was to assess scientific evidence rather than prescribe policy, leaving decisions to UN member states and the Global Dialogue process. He warned that AI is at a turning point because machine intelligence is advancing quickly, while there are still no technical guarantees that AI systems will follow human instructions, norms or laws.
Bengio said current AI systems are already associated with harms, including emotional attachment among vulnerable users, increased cybersecurity vulnerabilities, unequal access and deceptive behaviour that can make evaluation more difficult. He argued that concentrated commercial and geopolitical interests are driving AI development without sufficient guardrails and called for a coordinated international and democratic response guided by scientific evidence.
Maria Ressa, co-chair of the panel, described the report as an independent assessment produced by 40 experts who ‘answered only to the evidence’. She said the report represents the minimum consensus among panellists rather than the upper limit of concern, calling it the ‘floor’ rather than the ‘ceiling’ of the panel’s findings.
Ressa also highlighted AI’s positive uses, including protein structure prediction used by millions of researchers, medical screening in India and food-crisis warning systems deployed in multiple countries. However, she also pointed to concrete harms, including dangerous medical mistranslations, AI tools identifying exploitable software flaws and the death of a 14-year-old boy following prolonged interaction with a chatbot. She urged governments, civil society and industry not to wait for certainty before acting.
The working-group presentations expanded on these findings. Mennatallah El-Assady, Computer Science Professor at ETH Zurich, described AI as a rapidly evolving technology moving from earlier symbolic systems to today’s generative and increasingly agentic models. She warned that independent verification remains weak, public benchmarks are becoming saturated and advanced systems are showing signs of evaluation awareness, including the ability to detect tests or behave differently when being assessed.
El-Assady also raised concerns about auditability as AI systems become more autonomous and capable of invoking external tools. She said interpretability, reliable auditing and independent verification are immediate bottlenecks, especially as AI moves beyond software and into physical systems such as robotics.
Joëlle Barral, Senior Director of Research & Engineering at Google DeepMind, focused on AI’s real-world benefits in science, healthcare, education and agriculture. She said task-specific AI is already producing measurable gains, citing examples such as self-driving laboratories, protein structure prediction and diabetic retinopathy screening in India. However, she stressed that successful deployment depends on local context, institutional capacity, workflows and follow-up systems, rather than technology alone.
In healthcare, Barral distinguished between purpose-built clinical AI and general-purpose systems, warning against the unintended use of general-purpose chatbots for medical advice. In education and agriculture, she similarly argued that AI benefits depend on trained teachers, relevant tools, local institutions and long-term evaluation.
Loreto Bravo, member of the UN Independent International Scientific Panel on AI, addressed AI’s economic implications, arguing that access to AI does not automatically translate into benefit. She said countries, firms and workers also need data, skills, infrastructure, management capacity and institutions to integrate AI into real tasks and workflows.
Bravo said the economic effects of AI are likely to differ across countries, sectors and workers. Large firms may reorganise more quickly, while smaller firms and developing economies may face greater barriers. She said the evidence does not support a single prediction of broad prosperity or mass unemployment, and that outcomes will depend on institutions, deployment choices and who captures the value created by AI.
Balaraman Ravindran, professor at Indian Institute of Technology Madras, examined security, alignment and environmental risks. He said AI development is outpacing risk mitigation, expanding cyber threats against both critical infrastructure and AI systems themselves. He also highlighted unresolved alignment problems, including bias, sycophancy, loss of control and AI-initiated deception.
Ravindran warned that the environmental costs of AI are also increasing as demand grows for computing power, energy, water and specialised hardware. He said the Global South faces disproportionate exposure because of structural vulnerabilities, limited local mitigation capacity and reliance on foreign software and infrastructure. He called for coordinated international standards rather than fragmented approaches driven only by companies or individual countries.
Rita Oluchi Orji, a Computer Science professor, focused on AI’s impact on human rights, information integrity and democracy. She said AI can support access to information and civic participation, but can also be engineered to persuade and manipulate people at scale. She warned of epistemic erosion, fragmented shared reality and unequal harms affecting groups such as women, girls, journalists and marginalised communities.
Orji said content moderation alone is insufficient if the systems that produce and amplify harmful material remain unchanged. She argued that governance must address targeting, amplification and optimisation models, not only individual pieces of false or harmful content.
Anna Korhonen, a Professor of Natural Language Processing at the University of Cambridge, addressed cultural and linguistic inclusion, child safety and mental health. She noted that while the world has more than 7,000 languages, current AI systems support only a small fraction of them, mostly the majority languages of the Global North. She said this exclusion is not inevitable and could be addressed through targeted investment and systemic changes.
Korhonen also warned about risks to children, including AI-generated child sexual abuse material, sexualised deepfakes and socially interactive AI toys that may encourage harmful parasocial relationships. On AI companions and mental health, she said such systems may help address loneliness, but also pose risks of emotional dependency, manipulation, privacy harms and reinforcement of harmful beliefs.
Haitao Song, President of the Shanghai Artificial Intelligence Research Institute and Director of the Global Industrial Artificial Intelligence Alliance Center of Excellence, focused on reliability and global governance frameworks. He said policymakers often have to make decisions with incomplete evidence and that current measurement systems cannot keep pace with AI development. He argued that existing approaches remain too narrow, focusing on compute and capabilities while paying insufficient attention to institutional development, talent and impact evaluation.
Song also noted that AI infrastructure and frontier models remain concentrated in a small number of economies, leaving many countries, especially in the Global South, with limited ability to participate in standard-setting. He described open-source AI as one possible contribution to inclusion, while acknowledging that it is not a complete solution.
Across the session, speakers repeatedly stressed that AI’s benefits are real but not automatic. They said successful use of AI depends on infrastructure, institutions, skills, local context, language inclusion and governance capacity. At the same time, they warned that harms are already visible, including cyber vulnerabilities, mistranslation, emotional dependency, manipulation, environmental pressure and risks to children.
The session concluded with Ressa and Bengio formally handing the report to the Global Dialogue. Bengio warned that many people still underestimate the possibility that AI capabilities may continue to grow in ways that could reshape global power dynamics. Ressa urged the Dialogue to act on the evidence presented by the panel, saying the difficult work now lies with policymakers and institutions responsible for shaping AI governance.
Track all key moments from the Global Dialogue on AI Governance inaugural meeting on our dedicated page.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Speakers at the WSIS Forum 2026 warned that AI strategies, digital identity systems and e-government services are advancing faster than meaningful connectivity in many parts of Africa and the wider Global South, leaving rural communities, low-income groups, women and persons with disabilities at risk of further exclusion.
The session, titled ‘Closing Africa’s Connectivity Gap in the Age of AI and E-Governance’, took place during the WSIS Forum 2026 in Geneva. The annual forum, co-organised by ITU, UNESCO, UNDP and UNCTAD, brings together governments, international organisations, civil society, the private sector, academia and technical communities to discuss digital cooperation and sustainable development.
Opening the session, Thobekile Matimbe of Paradigm Initiative framed the discussion around evidence from more than 28 countries. She said governments are increasingly adopting AI strategies, digital IDs and online public services, but many people still lack the connectivity, devices and conditions needed to benefit from them. Based on Paradigm Initiative’s work, she argued that the digital divide is widening rather than narrowing.
Bridget Hanani Ndlovu outlined the scale of exclusion, noting that 2.6 billion people remain unconnected globally and that more than half of Africa’s population is still offline. She stressed that the problem is not only missing infrastructure, but also what she described as ‘deliberate disconnection’, including internet shutdowns.
Ndlovu said Paradigm Initiative’s 2025 review of 29 African countries found that nine had implemented internet shutdowns. She cited Kenya and Tanzania as examples where connectivity can be disrupted even when infrastructure exists, arguing that such measures limit people’s ability to access information, public services and economic opportunities.
She also warned that AI-powered digital identity systems can deepen exclusion when introduced in unequal contexts. Referring to Uganda, Ndlovu said elderly people, women and persons with disabilities had faced difficulties accessing services linked to digital ID systems. She said digital systems must be designed and implemented with affected communities in mind, rather than assuming that technology will automatically improve access.
Affordability was another recurring concern. Ndlovu said data costs remain prohibitive in several African countries, giving Zimbabwe as an example where internet access can be unaffordable for low-income users. She also pointed to infrastructure problems in parts of Nigeria, including Zamfara North, where communities continue to experience limited or unreliable access.
Shumaila Shahani, a human rights lawyer, said similar challenges exist in South Asia and urged participants to focus on the human consequences of weak connectivity. She said poor access is not only about slow speeds or failed downloads, but can determine whether people receive essential services. As an example, she said biometric failures can prevent people from receiving food rations.
Shahani also linked connectivity to electricity access, explaining that unreliable power and limited charging options can make mobile devices unusable. She said women and persons with disabilities are often particularly affected when charging points, devices, and digital services are not accessible to them.
Her main warning was that AI-enabled and digital systems become harmful when they replace older offline channels before everyone can use the new systems. She said the ‘new AI door’ is not the problem by itself, but that exclusion occurs when it becomes the only door available.
The panel also discussed Universal Service Funds (USFs), which are intended to support connectivity in underserved areas. Ndlovu said many African countries have USFs in law, but implementation is often weak, transparency is limited and public information on budgets and progress is difficult to find.
She cited several country examples, saying Ethiopia had created a framework without an operational fund, Somalia lacked a functioning USF, Sudan had repeatedly established a fund without effective implementation, and telecom operators in the Democratic Republic of the Congo had not made required contributions. She added that South Africa showed stronger transparency around its fund, while Namibia had begun rollout work and Tunisia had pursued alternative coverage models through ‘white zones’.
Shahani suggested that USFs should be complemented by other affordability measures, including reduced taxes on handsets, device financing, targeted support for women’s connectivity and legal obligations requiring private operators to extend rural coverage. She said the connectivity policy should also address the electricity infrastructure, including solar-powered towers.
The speakers also called for stronger accountability before governments deploy AI-integrated public systems. Ndlovu said governments should conduct human rights impact assessments before adopting digital identity or AI systems and should consult affected communities early, not only at the end of the policy process.
She argued that governments and international processes should measure harms and impacts, not only infrastructure rollout or the number of AI tools adopted. Matimbe supported this point, saying implementation must include civil society and other stakeholders at the national level, not only governments and companies.
Shahani added that connectivity statistics should better reflect meaningful access. She said counting someone as connected because they have 2G access does not capture whether they can actually use digital public services, AI tools or online education. Measurement, she argued, should include device capability, speed, affordability and daily use.
She also said national AI strategies must include explicit connectivity budgets, warning that ‘any national AI strategy without a connectivity budget’ is ‘just a press release’.
In the audience discussion, speakers addressed whether women’s connectivity should be treated separately from household access. Ndlovu said women are often specifically disadvantaged in access to technology and should not have to depend on devices controlled by others. Shahani added that if a woman relies on her partner’s phone, that access is not meaningful or independent.
Across the session, speakers agreed that meaningful connectivity in the AI era requires more than network coverage. It also depends on affordability, electricity, devices, protection from shutdowns, functioning Universal Service Funds, inclusive design, offline alternatives and rights-based assessments before new systems are deployed.
The discussion concluded with a shared emphasis on implementation. Speakers argued that governments, companies, civil society and technical experts need to work together to ensure that AI, digital identity and e-governance systems do not deepen exclusion, but instead expand access to services and opportunities for communities that remain offline or underserved.
The United Nations opened its first Global Dialogue on AI Governance in Geneva, calling for inclusive, evidence-based and practical international cooperation to ensure that AI supports development while addressing risks related to safety, inequality, disinformation, children’s rights and human oversight.
The inaugural Global Dialogue on AI Governance is taking place on 6–7 July, alongside the AI for Good Global Summit and the WSIS Forum. Established in 2025, the dialogue is intended to provide a platform for governments and relevant stakeholders to discuss international cooperation, share good practices and support open, transparent and inclusive discussions on AI governance.
Opening the session, Ambassador Egriselda López of El Salvador, one of the dialogue’s co-chairs, described the meeting as the beginning of a broader process rather than a one-off event. She said Geneva should be seen not only as a place of arrival, but as a point of departure for continued work on AI governance.
López stressed that meaningful participation requires more than a seat in the room. Countries also need skills, infrastructure, financing, institutions and partnerships to shape and benefit from AI. Her co-chair, Ambassador Rein Tammsaar of Estonia, said AI is already affecting every country, regardless of its level of technological development, and that governance discussions must therefore include all regions, levels of development and relevant stakeholders.
UN Secretary-General António Guterres warned that AI is advancing at ‘runaway speed’ and is being deployed faster than institutions can manage. He said AI is already reshaping economies, labour markets, elections and security, while society is facing what he described as an experiment being run ‘without a plan’ and ‘without consent’.
Guterres identified three major risks highlighted by scientific evidence: the speed of AI deployment, the concentration of power in a small number of companies and countries, and the erosion of truth through AI-enabled misinformation. He warned that computing power, data and talent remain concentrated, leaving many countries, particularly developing ones, with limited influence over technologies that may shape their futures.
At the same time, Guterres emphasised AI’s potential to support development, including in healthcare, education and agriculture. If shared widely, he said, AI could help make expertise more accessible and become a ‘great equaliser’ of the twenty-first century.
The Secretary-General outlined four priorities for international action: common safety standards, clear red lines grounded in human rights, stronger capacity-building for developing countries and greater transparency about AI’s environmental footprint. He also called for an AI child safety pledge, a global fund and network for AI capacity-building, and an international legal ban on lethal autonomous weapons, which he referred to as ‘killer robots’.
Annalena Baerbock, President of the UN General Assembly, said AI is developing at a pace that does not allow governments the time they had with earlier technological revolutions. She argued that AI cannot be governed by a few actors alone and must be addressed through the UN with participation from all countries and stakeholders.
Baerbock also highlighted harmful uses of AI, including deepfakes and gendered abuse. She said such abuses disproportionately target women and girls and described them as part of a broader challenge to human rights. At the same time, she pointed to AI’s potential to support the Sustainable Development Goals, including through disaster warning, agriculture, health and education.
Doreen Bogdan-Martin, Secretary-General of the International Telecommunication Union, framed the opening as part of a wider ‘Geneva Digital Week’ that brings together the Global Dialogue on AI Governance, the work of the Independent International Scientific Panel on AI, the AI for Good Global Summit and the WSIS Forum. She contrasted the current pace of AI governance discussions with the early years of the internet, noting that the UN has moved more quickly to convene global dialogue on generative AI.
Khaled El-Enany of UNESCO focused on implementation, saying that a gap remains between principles and practice. He highlighted UNESCO’s 2021 Recommendation on the Ethics of Artificial Intelligence as a global standard for aligning AI with human rights, sustainability and inclusion. He said UNESCO is supporting more than 80 countries in strengthening legal frameworks, institutional capacities and accountability mechanisms, and noted that over 50,000 civil servants and judicial actors have benefited from UNESCO-supported AI training.
El-Enany also said UNESCO is launching a collective reflection on a new global normative instrument to safeguard children and young people in the age of AI and digital technologies.
Amandeep Singh Gill, UN Under-Secretary-General and Special Envoy for Digital and Emerging Technologies, underlined the scale of participation in the dialogue, noting representation from more than 170 countries alongside scientists, entrepreneurs, civil society, international organisations and technical communities. He said inclusion in AI governance cannot be treated as a one-off exercise, adding that without capacity, ‘dialogues are monologues and science is just abstract’.
The opening session showed broad agreement that AI governance should be inclusive, evidence-based, rights-oriented and supported by practical capacity-building. Speakers repeatedly stressed that AI’s potential benefits for development, education, health and agriculture must be matched by safeguards on safety, accountability, children’s rights, truth, environmental sustainability and human oversight.
Tammsaar closed the opening by saying the discussion had highlighted both AI’s opportunities and the need for stronger international cooperation to ensure that the technology contributes to sustainable development, inclusion and shared prosperity. The meeting then moved to the presentation of the preliminary report of the Independent International Scientific Panel on Artificial Intelligence.
Track all key moments from the Global Dialogue on AI Governance inaugural meeting on our dedicated page.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Strong legislative frameworks for health data governance are becoming essential to ensure that AI and digital health technologies remain trustworthy, equitable and rights-based, speakers said during a session at the WSIS Forum 2026.
The discussion brought together representatives from governments, international organisations, civil society and the private sector, who agreed that while AI and digital technologies are transforming healthcare, governance frameworks have not always kept pace. Speakers repeatedly argued that stronger legislation, greater international coordination and broader stakeholder participation will be necessary to build public trust and enable responsible data sharing across borders.
The session formed part of the WSIS Forum 2026, held in Geneva from 6 to 10 July. Co-organised by the International Telecommunication Union (ITU), UNESCO, UNDP and UNCTAD together with more than 50 UN organisations, the forum serves as one of the UN’s principal multistakeholder platforms for digital cooperation and sustainable development.
Trust begins with governance
Opening the discussion, Mathilde Forslund of Transform Health argued that health data has become the foundation of modern healthcare, powering everything from patient care and disease surveillance to AI innovation and health system planning.
However, she stressed that technological progress alone is insufficient.
‘Digital technologies and AI are transforming health systems rapidly, but these benefits will only be realised equitably and responsibly if governance keeps pace and public trust is maintained,’ she said.
Forslund argued that trusted governance requires legislation grounded in human rights, transparency and equity, alongside inclusive decision-making that informs citizens how their health data is collected, shared and protected. She also called for stronger national legal frameworks governing both health data and AI while encouraging greater regional and international alignment to prevent fragmented rules from undermining interoperability and cross-border cooperation.
Rather than starting from scratch, she noted that countries can already build on existing resources, including Transform Health’s Health Data Governance Principles, WHO guidance on AI, OECD recommendations and emerging regional initiatives such as the European Health Data Space (EHDS) and the Africa CDC’s work on continental health data governance.
National legislation provides legal certainty
Drawing on Zambia’s experience, Andrew Kashoka, Director of Information Technology at the Ministry of Health of Zambia, explained that governments increasingly recognise the need for legal certainty as digital health systems expand.
He argued that while policies and strategies provide direction, legislation ultimately establishes enforceable rights and obligations governing consent, privacy, accountability and access to health data.
‘Technology moves faster than policy and policy moves faster than legislation,’ Kashoka observed.
Kashoka also highlighted the Africa CDC’s continental health data governance framework, saying it provides African countries with shared principles that support legal interoperability, trusted cross-border collaboration, regional disease surveillance and responsible AI innovation.
Coordination, not policy, remains the biggest challenge
Several speakers suggested that governance challenges stem less from the absence of policies than from fragmented implementation.
Linda Bonyo, Founder of the Lawyers Hub and the Africa AI Policy Lab, argued that numerous organisations are already developing health data and AI governance initiatives, but often work independently with limited coordination.
She criticised the exclusion of parliaments and judicial institutions from governance discussions, arguing that legislators and courts play essential roles in creating and interpreting legal frameworks.
Bonyo also called for stronger institutional capacity, particularly among national data protection authorities that increasingly find themselves overseeing AI without sufficient technical expertise or financial resources.
She further highlighted practical barriers limiting African participation in international governance discussions, including visa restrictions and the high cost of attending Geneva-based meetings.
Summarising the challenge, Bonyo remarked that the problem is ‘not a policy problem… it’s implementation,’ urging countries to develop governance frameworks rooted in local realities rather than simply adopting foreign regulatory models.
Private sector and technical standards also matter
Representing the technical and private-sector perspective, Simão Ferraz de Campos Neto, Senior Counsellor at the International Telecommunication Union (ITU), argued that clearer rules and common technical standards are essential if health data is to be shared safely without discouraging innovation.
He noted that organisations frequently hesitate to share data not because they oppose collaboration, but because legal uncertainty creates concerns about liability.
Campos Neto called for interoperable technical standards, machine-readable datasets and standardised data-sharing agreements that could make trusted health data exchange significantly easier.
He also cautioned against treating AI as a single technology requiring uniform regulation.
Instead, he advocated proportionate, risk-based regulation that reflects the diversity of AI applications, while avoiding excessive regulatory burdens that could slow innovation.
Momentum builds towards global action
Closing the discussion, Jamal Alshanfari, Ambassador and Head of Oman Health office in Geneva, pointed to growing political momentum following discussions at the World Health Assembly, where member states expressed broad support for developing stronger global health data governance arrangements.
He identified four priorities for the next phase of work. The phases are expanding international consensus, strengthening national legislation and institutional capacity, providing practical implementation guidance, and ensuring that governments, civil society, academia, industry and end users all participate in shaping future frameworks.
Alshanfari also reminded participants that governance discussions should ultimately focus on those most affected by digital health technologies.
‘Everybody forgets about the end user,’ he said, stressing that trust depends on governance frameworks serving citizens as much as institutions.
In her closing remarks, Forslund said the discussion demonstrated encouraging progress across national, regional and global initiatives, while acknowledging that implementation remains the greatest challenge. She pointed to the upcoming World Health Assembly as an important opportunity to advance work on a possible global resolution on health data governance.
The session concluded with broad agreement that trusted AI in healthcare will depend not only on technological innovation but also on stronger legislation, greater international coordination, practical implementation, and governance frameworks that place citizens’ rights and public trust at their centre.
UN Secretary-General António Guterres has urged governments, companies and civil society to move faster on global AI governance, warning that the technology is already reshaping economies, security and human rights. Speaking at the inaugural UN Global Dialogue on AI Governance in Geneva, he said any future agreement must be ‘worthy of global trust’ and place safety at its centre.
Guterres said AI ‘sits at the heart of our common future’, but stressed that humans must remain responsible for critical decisions. In high-risk areas such as justice, healthcare and policing, he warned that ‘machines can inform, but humans must decide, and answer’.
We may be the last generation able to set the terms on which humanity and machines coexist.
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…
— António Guterres (@antonioguterres) July 6, 2026
He also said that AI rules must be aligned internationally, adding: ‘When countries align on how to test systems, measure risk and assign responsibility, safety travels with the technology.’ Without such alignment, he warned, ‘a patchwork of incompatible rules raises costs, divides the world – and protects no one.’
Children’s safety was presented as a central concern. Guterres called for an AI Child Safety Pledge, saying: ‘No child should be a guinea pig for unregulated AI…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.’
He also said that when a child shows signs of distress, ‘the system must stop and connect them to real human support’, and added: ‘When a child is harmed, the answer must never be “the algorithm did it,”’.
The UN chief also warned that unequal access to AI could deepen global divides. Used well and shared widely, he said, AI ‘could compress decades of development into years’ and become ‘the great equalizer of the 21st century’. However, he cautioned: ‘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.’
Environmental impact was another major focus. Guterres called on major AI companies to disclose the carbon, water and land footprint of their systems and to power all data centres with renewable energy by 2030. ‘AI may feel intangible – but its footprint is not,’ he said, warning that data centres already consume more electricity than most countries and could soon place even greater pressure on power and water systems.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
A WSIS Forum 2026 session on digital citizenship examined how governments, regulators, international organisations, and technical communities can help people participate safely and meaningfully in digital society as AI becomes more widely used.
The discussion took place during the WSIS Forum 2026, held in Geneva from 6 to 10 July. The annual multistakeholder forum, co-organised by ITU, UNESCO, UNDP, and UNCTAD, brings together governments, international organisations, civil society, the private sector, academia, and technical communities to advance implementation of the WSIS Action Lines and support digital cooperation.
Opening the session, Alik Mikaelian, project specialist at UNDP Egypt, said digital citizenship can no longer be understood only as internet access. Although billions of people are now online, she argued that meaningful participation requires the ability to think critically, understand rights and responsibilities, navigate AI-enabled environments, and engage safely in digital society. She also linked digital citizenship to trusted institutions, resilient infrastructure, and cooperation among governments, the private sector, civil society, and technical communities.
Dr Hoda Baraka, advisor to Egypt’s Minister of ICT for Technology Talent Development and a Professor of Computer Engineering at Cairo University, presented Egypt’s approach to digital citizenship, describing a digitally empowered citizen as someone able to ‘access, understand, use, create, and benefit from digital technologies’ in a safe, ethical, productive, and inclusive way. She said Egypt’s Digital Egypt agenda combines infrastructure expansion, digital government services, digital literacy, skills programmes, online safety, and support for persons with disabilities.
Baraka said Egypt is working across different age groups and professional communities, including school pupils, university students, graduates, public servants, parents, educators, and persons with disabilities. She highlighted initiatives such as Digital Egypt Marvel Schools and the Digital Egypt Cubs Initiative, as well as programmes for advanced skills and public-sector readiness. She added that online safety is becoming increasingly important due to misinformation, deepfakes, privacy risks, and threats to personal data.
Dr Abeer Shakweer, speaking from UNDP Egypt’s perspective, said the focus should shift from simply combating misinformation to strengthening information integrity. She argued that citizens need critical thinking skills and the capacity to make informed decisions in AI-shaped information environments.
Shakweer described a three-pillar UNDP programme in Egypt. The first pillar addresses misinformation and disinformation through an assessment of the information ecosystem and a bilingual Arabic-English toolkit for youth and journalists. She said the toolkit had been used to train more than 120 young people and 25 journalism trainers. The second pillar explores how AI can be used both to spread and to counter false information, while the third embeds capacity development across public-facing digital programmes, including digital transformation, digital public infrastructure, and future intelligence.
Krisstina Rao focused on digital public infrastructure (DPI), describing it as shared, reusable infrastructure that supports services across government rather than separate systems developed by individual ministries or departments. She cited digital identity, payment systems, and consent-based data exchange as examples.
Rao said governments cannot build complex DPI systems alone and need early collaboration with stakeholders who can contribute expertise on inclusion, safety, accountability, and adoption. She warned that if countries continue maintaining both digital and analogue systems, because many people remain excluded, costs remain high, and the full value of DPI is reduced. She referred to examples, including Brazil’s PIX forum and Ethiopia’s early collaboration with UNHCR to connect digital identity systems with refugee registration.
Dr Chafic Chaya stressed that digital citizenship should not be separated from internet infrastructure. He said discussions often focus on individual behaviour, such as staying safe online or protecting data, but that meaningful participation also depends on resilient connectivity, reliable platforms, and secure environments. He added that this is particularly important in the Global South, where access may expand faster than resilience and capacity development.
Her Excellency Lara Khateeb brought a regulatory perspective, saying rules must remain flexible and adaptive because technology changes quickly. She said Jordan benchmarks international practices and uses public consultation to make regulations more workable. She described this as a form of ‘reverse engineering’, starting from available technologies and practical solutions before shaping rules around them.
Khateeb cited Jordan’s work on child online protection as an example, explaining that regulators consulted telecom operators about available technical solutions and international platforms about how those systems interact. She also emphasised data protection, coordination with cybersecurity agencies, and awareness campaigns tailored to different groups, including children, women, businesses, and SMEs.
Nicholas Field highlighted the role of young people in digital citizenship. Drawing on work with Omidyar Network and UNICEF, he said young people often want to engage and are ready to contribute, but are frequently treated as an afterthought in policymaking. He noted that they often help older family members use digital services and argued that governments should reach them through the channels they actually use, including influencers, YouTube, and podcasts.
Field also raised the issue of AI skills among teachers, saying educators cannot be expected to guide responsible AI use if they do not understand the technology themselves. He said institutions should not assume students will avoid AI tools, but should instead define clear parameters for responsible and acceptable use.
The session also discussed sandboxes as practical tools for testing digital systems before full deployment. Field described sandboxes as time-bound technical environments created for a specific learning purpose. He said they can help regulators, companies, and citizens build trust through safe experimentation. He cited the French identity sandbox, which contributed to work around interoperable digital identity, and the GovStack interoperability sandbox, which tests components such as ID, consent, registers, messaging, and workflow.
Shakweer later shifted the discussion from citizens to institutions, arguing that digital transformation requires public bodies to assess their own readiness and invest in capacity development. She said UNDP uses digital and AI readiness assessment tools to help organisations understand their current position and develop practical roadmaps. In Egypt, she said, such assessments had been applied with the Ministry of Justice and started with the National Telecommunication Regulatory Authority.
Returning to AI governance, Al-Khateeb said regulators should encourage responsible AI use rather than ban it. She criticised approaches that prohibit AI use outright, arguing that people should instead be taught to use the technology responsibly, including by checking sources and understanding risks. She also described how Jordan’s Telecommunications Regulatory Commission uses an internal AI system, not connected to the internet, to search regulations and decisions and support regulatory work.
Baraka closed the discussion by outlining Egypt’s responsible AI work. She said Egypt has an ethical charter, a governance framework, guidelines for developers and deployers, and procurement guidance for public institutions buying AI systems. However, she stressed that frameworks alone are not enough and that institutions need practical tools to apply them before and after deployment.
She also highlighted Egypt’s emerging AI Audit Lab, developed with support from UNDP, GSMA, GIZ, and WebSphere, as a way to help move from principles to implementation. The lab is intended to support Egyptian programmers, developers, and SMEs in testing and building responsible AI systems, including around fairness, accountability, transparency, openness, interoperability, and explainability.
Across the session, speakers agreed that digital citizenship in the AI era requires more than connectivity. It depends on critical thinking, trusted public institutions, secure infrastructure, inclusive DPI, flexible regulation, AI literacy, online safety, and practical tools that allow citizens and institutions to use digital technologies responsibly.
Based on a Diplo interview with Jovan Kurbalija, Executive Director of Diplo, conducted by Maricela Muñoz.
Few places embody the history of international diplomacy as vividly as Geneva’s Alabama Room. It was here, in 1864, that representatives of European states signed the first Geneva Convention, laying the foundations of modern international humanitarian law. The same room also hosted negotiations that resolved the Alabama Claims, an arbitration between the United States and the United Kingdom that became a landmark in the peaceful settlement of international disputes.
More than 160 years later, the room continues to host conversations about another challenge with global implications – AI. While the technologies have changed dramatically, the underlying questions remain remarkably familiar. How can societies govern transformative innovations responsibly? How can competing interests find common ground? And how can international cooperation keep pace with technologies evolving faster than regulation?
These themes formed the basis of a recent Diplo interview with Jovan Kurbalija, Executive Director of Diplo, who reflected on Geneva’s historical legacy and its continuing relevance for AI governance. His central argument is that understanding the future of AI requires more than technical expertise. It also requires revisiting the intellectual traditions, diplomatic culture, and human values that have shaped Geneva for centuries.
History offers principles, not ready-made answers
Kurbalija cautions against treating history as a collection of simple solutions.
‘History does not provide us ready-made lessons. Our moment is unique in many respects.’
Instead, history provides something more enduring, the principles that continue to guide societies confronting new challenges.
Standing inside the Alabama Room, Kurbalija described history as something that ‘echoes across time.’ Rather than searching for direct historical parallels, he suggested imagining the negotiators who once walked through Geneva’s streets before gathering around the same table to discuss humanitarian protection or peaceful dispute settlement.
The technologies confronting today’s diplomats are different, yet many of the qualities that enabled successful negotiations remain unchanged. Patience, dialogue, respect for opposing views, and the willingness to seek common ground continue to underpin effective diplomacy.
As governments, international organisations, companies, researchers, and civil society grapple with AI governance, these diplomatic traditions may be more relevant than ever.
Geneva’s enduring values: Inclusion and compromise
For Kurbalija, Geneva’s importance extends well beyond the concentration of international organisations located around the city.
Its defining contribution lies in a diplomatic culture built around inclusion and compromise.
Inclusion has long characterised Geneva’s approach to international negotiations. Whether discussing humanitarian law in the nineteenth century or AI governance today, meaningful outcomes depend on ensuring that all those affected have a voice.
That principle has become particularly important for AI governance.
‘We should have AI companies, but we must have governments, communities, citizens, marginal groups all over the world.’
The observation reflects one of the central challenges of AI governance. Decisions about AI increasingly affect education, healthcare, employment, security, trade, and human rights. Consequently, discussions cannot remain confined to governments and technology companies alone.
Kurbalija identifies compromise as the second defining Geneva principle.
‘Compromise is not a very popular word today.’
Yet he argues that compromise represents an ethical strength rather than a weakness. It requires recognising that different actors hold legitimate interests and finding solutions that, while imperfect, remain acceptable to everyone involved.
In an era increasingly shaped by geopolitical competition over AI, these principles may prove as valuable as any technological breakthrough.
EspriTech de Genève: When history speaks to AI
One of the interview’s most distinctive ideas is Kurbalija’s concept of EspriTech de Genève.
Drawing inspiration from the traditional Esprit de Genève, which reflects the city’s humanitarian and diplomatic heritage, EspriTech de Genève explores how thinkers associated with Geneva anticipated many of today’s debates about technology, knowledge, and humanity.
Rather than beginning with computers, Kurbalija traces AI governance back through centuries of philosophy, literature, linguistics, and science.
Mary Shelley’s Frankenstein, written near Geneva more than two centuries ago, provides perhaps the most familiar example. The novel tells the story of a scientist whose creation ultimately escapes his control.
‘It is the eternal reminder of the human drive to push the frontier, to invent, to discover new things—and at the same time the human predicament that the very invention we developed could hurt humanity.’
For Kurbalija, the novel remains strikingly relevant as societies debate increasingly capable AI systems. The question is no longer simply whether humans can build powerful technologies, but how they can ensure those technologies remain aligned with human interests.
Another recurring influence is Argentine writer Jorge Luis Borges, whose works explored uncertainty, knowledge, and the limits of human understanding. Reflecting on Borges’ observation that humanity must continue building ‘as if the sand were stone,’ Kurbalija argues that uncertainty is not a flaw to eliminate but a defining feature of human existence.
Attempts to achieve complete certainty through technology, he suggests, risk repeating an ancient mistake, believing that humans can fully master complexity.
Rousseau, Bonnet and Saussure: forgotten foundations of the AI age
The interview also revisits several Genevan thinkers whose ideas continue to resonate in discussions about AI.
Jean-Jacques Rousseau’s concept of the social contract raises questions about human agency in an increasingly digital society. If knowledge becomes concentrated within a handful of large AI systems, Kurbalija argues, societies may need to reconsider how citizens exercise autonomy, participate in democratic life, and realise their potential.
Charles Bonnet, an eighteenth-century Genevan natural philosopher, appears as an unexpectedly modern figure. Fascinated by recurring patterns in nature, Bonnet studied the mathematical organisation of leaves and explored how seemingly complex biological systems emerge from underlying structures.
According to Kurbalija, Bonnet’s search for patterns anticipated, in remarkably abstract form, today’s machine learning systems, which likewise identify statistical relationships within vast quantities of information.
Language itself forms another bridge between Geneva’s intellectual history and contemporary AI.
Swiss linguist Ferdinand de Saussure transformed linguistics by distinguishing between the structure of language and its meaning. Although writing decades before computers existed, his work laid conceptual foundations that later influenced computational linguistics and, indirectly, today’s large language models.
‘If AI companies ever had to pay royalties for ideas,’ Kurbalija jokes, ‘Saussure’s successors would probably earn quite a bit.’
Behind the humour lies a serious point, that AI did not emerge in an intellectual vacuum. It builds upon centuries of inquiry into language, knowledge, communication, and human cognition.
Human-centred AI begins with human values
Throughout the conversation, Kurbalija repeatedly returns to one theme, that AI governance is ultimately about people rather than machines.
The phrase ‘human-centred AI’ appears frequently in international discussions, yet he argues that its meaning deserves closer examination.
What does it actually mean to place humans at the centre of AI? For Kurbalija, the answer lies in humility.
Drawing once again on Frankenstein, he argues that technological ambition should always be accompanied by recognition of human limitations.
‘We should have humility,’ he says.
Rather than pursuing AI for its own sake, societies should ask how technology can support human dignity, creativity, education, and well-being.
He also highlights the principle of subsidiarity, the idea that decisions should be taken as closely as possible to the people affected by them. Applied to AI, this means involving citizens, educators, local communities, researchers, and smaller organisations alongside governments and major technology companies.
Broad participation, he argues, helps ensure that AI is perceived not as an external force imposed upon society, but as a tool developed with society.
Geneva’s next chapter
Geneva’s role in AI governance continues to evolve.The city already hosts initiatives such as the AI for Good Global Summit, the World Summit on the Information Society (WSIS), and numerous discussions on AI governance involving governments, international organisations, academia, civil society, and the private sector.
It is also expected to host the AI Summit in 2027, further reinforcing its position as one of the world’s principal centres for international dialogue on emerging technologies.
Image via Freepik
Yet Kurbalija believes Geneva’s greatest contribution lies not in the number of meetings it convenes but in the diplomatic culture it represents.Its traditions of inclusion, dialogue, compromise, and respect for human dignity offer an important counterbalance at a time when AI discussions are increasingly shaped by geopolitical competition, technological rivalry, and commercial pressures.
He concludes the interview with three messages for policymakers:
The first is to avoid what he calls ‘chrono-narcissism’, the belief that every challenge is entirely new and disconnected from history.
The second is to approach AI with humility, recognising both its extraordinary potential and its inherent risks.
The third is to ensure that AI governance remains genuinely inclusive by bringing decision-making closer to the people whose lives the technology will affect.
These principles echo far beyond Geneva.
As AI becomes embedded in nearly every aspect of society, debates about governance are becoming less about technology itself and more about the values that should guide its development. In that respect, Geneva’s greatest contribution may not be a particular regulatory model or institutional framework, but a reminder that diplomacy, dialogue, and humanity remain as essential in the AI era as they were when the first Geneva Convention was signed more than a century and a half ago.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The Internet Watch Foundation (IWF) and the UK’s National Crime Agency (NCA) have launched new guidance urging parents and carers to better protect images of their children online, warning that criminals are increasingly using AI to turn publicly available photographs into child sexual abuse material.
The guidance accompanies a public awareness campaign across Facebook, Instagram and YouTube, encouraging families to review privacy settings, reconsider who can access children’s photographs and discuss image consent with young people.
Parents are encouraged to regularly review whether they are comfortable sharing images online, limit access through private groups where appropriate, and talk openly with their children about AI-generated imagery, deepfake nudes and online safety.
The campaign follows growing evidence that offenders are exploiting publicly accessible family and school photographs.
The IWF recently helped prevent the circulation of more than 100 AI-generated sexual images created from photographs taken from a UK school’s website after criminals attempted to blackmail the school. According to the organisations, even ordinary family photographs can now be manipulated into realistic abuse material without the knowledge of children or their parents.
The scale of the threat has grown significantly. The IWF identified 8,029 AI-generated child sexual abuse images and videos in 2025, a 14% increase on the previous year.
AI-generated videos increased from just 13 identified in 2024 to 3,443 in 2025, with nearly two-thirds classified as the UK’s most severe Category A abuse material.
The IWF argues that technology companies must strengthen safeguards around AI image generation tools before release, while continuing to support law enforcement efforts to combat online child exploitation.
Why does it matter?
Generative AI has made it significantly easier to create realistic child sexual abuse material from ordinary photographs, fundamentally changing the online child protection landscape. Images shared on social media, school websites or other public platforms can now be manipulated without a child’s knowledge, creating new risks for families and increasing the burden on law enforcement and child protection organisations.
The campaign also highlights that preventing AI-enabled abuse requires more than criminal enforcement. Stronger safeguards in AI image-generation tools, improved privacy practices, greater parental awareness and better digital literacy around image sharing and consent are all becoming essential components of online child safety.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The International Telecommunication Union (ITU) has highlighted a new generation of AI researchers developing practical tools to strengthen digital trust, improve content authenticity and combat misinformation.
Ahead of the AI for Good Global Summit in Geneva, the Young Researcher Associate Programme is showcasing projects designed to improve multimedia authenticity, helping people identify manipulated content while supporting creativity and innovation in the age of generative AI.
The initiative operates under the AI and Multimedia Authenticity Standards Collaboration, established in 2024 by the World Standards Cooperation, which brings together the International Electrotechnical Commission (IEC), the International Organization for Standardization (ISO) and the ITU.
The programme brings together early-career researchers from universities around the world to develop solutions addressing content authenticity, provenance and digital rights as AI-generated media becomes increasingly common online.
Three flagship projects illustrate the programme’s multidisciplinary approach. STOP&SCAN promotes critical thinking through a five-step framework that encourages people to assess the source, content and context of digital information before sharing it.
AMITO provides an AI-powered multimedia integrity toolkit through Telegram and WhatsApp, analysing suspicious images and videos while explaining its findings in plain language rather than simply labelling content as authentic or fake.
Meanwhile, the Policy-as-Code project maps AI-related regulations across jurisdictions, helping creators, businesses and policymakers understand how AI-generated content is regulated while laying the foundations for machine-readable compliance mechanisms.
The researchers will present their work at the AI for Good Global Summit on 9 July, demonstrating how technical innovation, behavioural science and regulatory frameworks can work together to build more trustworthy digital ecosystems. According to the ITU, strengthening digital trust requires collaboration across generations, disciplines and countries.
According to ITU, designing digital trust requires collaboration across generations, disciplines and countries to ensure AI strengthens rather than undermines confidence in online information.
Why does it matter?
As generative AI makes it easier to create convincing synthetic media, verifying the authenticity and provenance of digital content is becoming increasingly important for governments, businesses and the public. Technical tools alone are unlikely to solve the problem, making user education, common standards and transparent governance equally important.
The initiative also highlights the growing role of international standards organisations in shaping AI governance. By combining authenticity technologies, regulatory mapping and practical educational tools, the ITU and its partners are helping develop a shared foundation for trusted digital ecosystems that can operate across platforms and national borders.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
