Global Law Forum

In my view, the first Global Dialogue on AI Governance would be successful if it creates a practical bridge between existing human rights obligations and the actual architecture of AI systems. Based on my book and current research on Global Digital Human Rights, I believe the key challenge is not only to regulate AI risks, but to ensure that AI systems are designed to realize human dignity, autonomy, equality, access to social benefits and human development. A successful Dialogue should therefore produce a shared implementation agenda, including: 1. recognition that human rights must be embedded into AI systems from the beginning; 2. support for developing a UN-level Declaration or Convention on Global Digital Human Rights; 3. a roadmap for Digital Human Rights Due Diligence for AI developers and deployers; 4. development of a Digital Rights-Compatibility Test for AI and digital systems; 5. guidance on transparency, accountability, human oversight, explanation and contestability; 6. support for public AI system registries and independent audits; 7. capacity-building that includes not only AI readiness, but also human rights readiness; 8. special safeguards for children, youth, vulnerable groups and future generations; 9. discussion of techno-legal standards translating human rights into technical and institutional requirements; 10. recognition of the need to redesign, restrict or suspend AI systems that create structural harm. The most important outcome would be a shift from protecting rights after harm has occurred to realizing rights through the architecture, governance, design, decision points and code of AI systems.

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

I selected "Protection and promotion of human rights" as the main priority because, in my view, human rights should be the normative core and main driver of global AI governance. Based on my book and current research on Global Digital Human Rights, I believe that AI governance should not be limited to safety, risk management or technical efficiency. These elements are important, but they should serve a deeper purpose: the realization of human dignity, autonomy, equality, access to social benefits, participation and human development in the digital age. I understand Global Digital Human Rights as opportunities belonging to every person from birth, aimed at preserving and developing human dignity and personality, enshrined in international law and domestic legislation, the implementation of which should enable access to social benefits through new technologies and without barriers created by those technologies. For this reason, transparency, accountability and human oversight are not separate goals, but implementation mechanisms. They are necessary to make human rights operational inside AI systems. A person affected by an AI system should be able to understand when AI is used, how a decision was made, who is responsible, how to challenge the decision and how to obtain an effective remedy. Safe, secure and trustworthy AI should also be understood through this rights-based lens. An AI system cannot be truly trustworthy if it is non-transparent, discriminatory, manipulative, unchallengeable or harmful to human autonomy. Therefore, the key task is to move from protecting rights after harm has occurred to realizing rights through the architecture, governance, design, decision points and code of AI systems. This could become the basis for a future UN Declaration or Convention on Global Digital Human Rights.

Yes. From my perspective, one cross-cutting issue is still not sufficiently visible: the gap between the speed of technological transformation and the absence of a shared legal language for the digital age. AI governance is often discussed through safety, risks, transparency, accountability and capacity-building. These themes are essential, but they do not fully address the deeper historical challenge: humanity is entering a digital environment where access to rights, opportunities and social benefits is increasingly mediated by AI systems, platforms, digital identity, data infrastructures and automated decisions. In my book, I describe this as a digital fork in history. We have technologies of the 21st century, but many of our legal and institutional frameworks still belong to an earlier era. The result is a growing imbalance: more technological power, but not necessarily more rights, dignity or human development. For this reason, I believe the missing cross-cutting issue is the need for a new global techno-legal language based on Global Digital Human Rights. This should include not only the protection of existing rights online, but also a clearer catalogue of digital human rights for AI, GovTech, Web 3.0, biometric systems, digital identity, data ecosystems and future technological environments. A UN-level Declaration on Global Digital Human Rights could become a starting point for this new social contract. It would help define not only what AI systems must not do, but also what they should enable: access to education, healthcare, public services, participation, information, creativity, development and the benefits of scientific progress. The emerging issue is therefore not only AI risk. It is whether global AI governance can become a framework for realizing human rights in the digital age, especially for children, young people, vulnerable groups and future generations.

In my sector of legal research, digital transformation and international digital governance, the main governance gap is the absence of a clear global framework for implementing human rights in digital systems. In my book, I describe this as a digital fork in history: technological development is accelerating, while legal and institutional frameworks remain fragmented and often reactive. This gap affects AI, GovTech, digital identity, biometric systems, Web 3.0, platforms, data infrastructures and automated decision-making. The most significant challenges include the lack of a universal catalogue of digital human rights, weak mechanisms for implementation, digital inequality, digital surveillance, digital militarization, algorithmic discrimination, excessive data collection, manipulation through digital platforms, and the vulnerability of children and young people in digital environments. Many systems still focus on efficiency, security or innovation, but not enough on dignity, autonomy, participation, access to social benefits and human development. In many countries and regions, another challenge is that AI readiness is developing faster than human rights readiness. States and companies may be ready to deploy AI technically, but not institutionally prepared to ensure transparency, accountability, human oversight, remedies and meaningful contestability. At the same time, these gaps create an important opportunity. AI governance can become a driver for a new global techno-legal framework based on Global Digital Human Rights. This could support the development of a UN-level Declaration, and eventually a Convention, on Global Digital Human Rights. Such a framework could help transform AI and digital systems from sources of risk into instruments for realizing rights: access to education, healthcare, public services, information, work, participation, creativity and the benefits of scientific progress. The key opportunity is to move from protecting rights after harm to implementing rights by architecture, governance, design, decision points and code.

The AI Dialogue can help move global AI governance from general principles to practical implementation. From my perspective, its main role should be to create a common international language for embedding human rights into AI systems and digital infrastructures. I believe this should be based on the concept of Global Digital Human Rights. The Dialogue could support: - recognition of Global Digital Human Rights as a guiding framework for AI governance; - development of a UN Declaration, and eventually a Convention, on Global Digital Human Rights; - transition from protecting rights after harm to realizing rights from the beginning of technological development; - translation of human rights into AI architecture, governance, design, decision points and code; - development of Digital Human Rights Due Diligence for AI systems; - creation of a Digital Rights-Compatibility Test for AI and digital systems; - safeguards for children, youth, vulnerable groups and future generations; - redesign or suspension of AI systems that create structural harm.

The AI Dialogue should build on the Global Digital Compact, but go further by creating a practical implementation layer for human rights in AI systems. From my perspective, the main missing element is a concrete framework of Global Digital Human Rights. The added value of the AI Dialogue could be: - to connect the Global Digital Compact with a future UN Declaration or Convention on Global Digital Human Rights; - to develop a catalogue of digital human rights for AI, GovTech, platforms, digital identity, biometric systems and automated decision-making; - to create techno-legal standards that translate human rights into technical and institutional requirements; - to introduce Digital Human Rights Due Diligence for AI developers, deployers and public authorities

Different stakeholders should contribute through concrete proposals, not only general statements. In my book, I emphasize the role of seven key stakeholder groups: the public sector, private sector, technical sector, legal sector, educational sector, media sector and the global community. For the AI Dialogue, each group should explain how it can help implement human rights in AI systems: - public sector: legal guarantees and accountability; - private sector: Digital Human Rights Due Diligence; - technical sector: safeguards in architecture, design and code; - legal sector: techno-legal standards; - education sector: AI and digital rights literacy; - media: public awareness and oversight; - global community: inclusive participation. The format should include short expert sessions, regional consultations, youth panels and practical working groups. The main output should be a roadmap toward a UN Declaration or Convention on Global Digital Human Rights and practical mechanisms for embedding rights into AI architecture, governance, design, decision points and code.

The most underrepresented voices are children, young people, vulnerable groups and communities affected by the digital divide. In my book, I argue that children and youth are not only future users of technology. They are already shaped by AI, platforms, digital education, data systems and online environments. Their identity, learning, creativity, socialization and development are increasingly influenced by digital systems. Other underrepresented voices include: - people without meaningful connectivity; - persons with disabilities; - linguistic minorities; - migrants and displaced people; - communities from countries with lower AI readiness; - users affected by automated public services; - civil society outside major AI powers. They should be included through youth consultations, regional dialogues, multilingual surveys, child-friendly formats, civil society panels and direct participation of people affected by automated decisions. AI governance should not be shaped only by those who build or regulate AI. It should include those whose lives are changed by it.

The AI Dialogue should not become only a sequence of formal speeches. It should use practical formats that test how human rights can actually be implemented in AI systems. Useful formats could include: - Digital Human Rights Labs; - youth and children's consultations; - regional dialogues; - case studies on AI in education, healthcare, public services and digital identity; - workshops on Digital Rights-Compatibility Tests; - sessions on techno-legal standards; - public repository of proposals and safeguards; - scenario exercises on digital surveillance, AI militarization, manipulation and exclusion. In my view, the most important format would be an implementation workshop on the future Declaration of Global Digital Human Rights. The Dialogue should help move from discussion to concrete tools: rights catalogues, safeguards, indicators, due diligence templates, audit mechanisms and redesign principles. The key task is to prototype how AI systems can realize human rights, not only how they can avoid harm.

Based on my book and current research, I would highlight several approaches that could offer concrete solutions for AI governance. 1. UN Declaration, and eventually a Convention, on Global Digital Human Rights. This could provide a common global framework for AI, GovTech, platforms, digital identity, biometric systems and automated decision-making. 2.techno-legal standards for AI systems. Human rights should be translated into technical and institutional requirements: data minimization, transparency, human oversight, contestability, auditability, safeguards for children and remedies. 3.Digital Human Rights Due Diligence. AI developers, deployers and public authorities should assess not only technical risks, but also how AI affects dignity, autonomy, equality, privacy, access to social benefits and human development. 4. rights-based digital public services. GovTech and AI in the public sector should be designed around the realization of rights: access to education, healthcare, social protection, participation, information and justice. 5. child- and youth-centred AI governance. Children and young people should not be treated only as vulnerable users. They should be recognized as key stakeholders whose identity, education, creativity and development are shaped by digital systems. These approaches reflect one core idea: AI governance should move from protecting rights after harm to implementing rights through architecture, governance, design, decision points and code.