Centre for AI & Digital Humanism - Digihumanism

To achieve the goal of the Global Dialogue of "becoming a truly global platform involving Governments and all relevant stakeholders, to discuss international cooperation, share best practices and lessons learned, and to facilitate open, transparent and inclusive discussions on artificial intelligence governance". (Res. 79/325) To come up with a common declaration in Geneva on current challenges and collaboration roadmap to address them, possibly through pledges from governments, tech companies, civil society, academia...

Interoperability of governance approaches;Safe, secure and trustworthy AI

- safe, secure and trustworthy AI is a requirement to achieve true innovation for the benefit of humanity. It is at the heart of both objective 3 and objective 5 of the Global Digital Compact. So the protection and promotion of human rights is actually a sub-category. Understanding and mapping the social, economic, ethical, cultural, linguistic and technical implications of AI is the first step in ensuring proper AI governance. Transparency, accountability and human oversight are the micro-governance means to achieve safe, secure and trustworthy AI. - interoperability of macro-governance approaches is key. The Global Dialogue has a crucial role to play by fostering global "AI cultural diplomacy" and multilateralism. This reflects Digihumanism's priorities since it aims to ensure a humanistic governance of AI. Since the spread of AI is global and touches upon all aspects of our lives, AI governance should be global as well in order to be effective.

#NAME?

The lack of an inclusive platform that engages all relevant stakeholders, combined with the ineffective representation of civil society in policy-making processes, directly affects our work and the sectors we engage with. These governance gaps result in concrete problems: Policies are developed without meaningful input from civil society, including academia, leading to AI regulations that, if they exist, prioritize short term technical and commercial interests over eg fundamental rights, which is detrimental to achieve AI innovation for the good of humanity. The lack of transparency and inclusiveness of decision-making processes means that human-centered concerns — such as algorithmic discrimination, privacy, and accountability — are often overlooked. Well-intentioned policies may thus fail to unleash AI benefits for Good by not addressing the risks associated with AI. The fast pace of AI development outpaces policy responses, and without continuous multistakeholder dialogue, governance frameworks risk becoming obsolete or ineffective. The opportunity, however, is that the Global Dialogue can fill this gap by establishing a permanent, inclusive platform where civil society and academia have an equal opportunity to participate alongside the public and the private sectors. With regard to "safe, secure and trustworthy AI", incidents detrimental to fundamental rights emerge on a regular basis in every parts of the world. Whatever the chosen path to regulate AI (hard law or agile governance), the minimum common denominator is the establishment of a governance framework along the lines of "map, measure, mitigate, govern"(digihumanism's own classification). There is thus the need to agree on the key steps and principles of this governance framework and nourish this framework through the Global Dialogue as well as the work of the Independent International Scientific Panel. With regard to AI agents, the risk is for these AI agents to become uncontrollable and/or make mistakes on behalf of individuals or governments with possibly dire consequences. There is a need to reassert the principle of human control and there is a need for humans to know when they interact with AI agents. Relatedly deep fakes have become a real challenge over the world. The issue of the integrity of the information space is particularly visible in relation to online manipulation in electoral context. Digihumanism has conducted several studies in this regard (eg in Romania or Germany) and has uncovered astroturfing modus operandi, with or without foreign interference. Without AI literacy, there is no way human control can be ensured on AI.

playing the part of a transparent, inclusive and continuous platform for AI cultural diplomacy: develop shared language of AI governance. identify trends, opportunities, risks and best practices towards shared diagnosis and solutions to ensure AI for Good Devrlop partnerships and pool human, institutional and financial resources together to address Ai challenges and opportunities and reduce the digital divide.

The Global Dialogue could serve as a coordination hub for UN activities in the field. Accordingly, the UN General Assembly could adopt a resolution calling other UN bodies to update and inform regularly the Dialogue about their work. The AI Dialogue could be further institutionalized with a secretariat to diffuse information about UN, States, civil society, academic or tech initiatives. the European Union Council of Europe CDNET Organisation of American States African Commission on Human and Peoples' rights

- Format of the AI dialogue: hub for transparent and inclusive collaboration - Structure: thematic sections + 1 general assembly + a secretariat - common sessions with different stakeholders, including states, academia, CSOs, tech companies - opportunities for dialogue and networking also among entities from the same stakeholder group

Balanced geographical representation: Most AI governance debates are dominated by Western countries, the EU, and the US. Perspectives from Africa, Latin America, are often disregarded, despite the global impact of AI systems. Indigenous communities: Indigenous perspectives on data sovereignty, collective rights, and traditional knowledge are almost entirely absent from mainstream AI governance frameworks. Vulnerable and marginalized groups: Persons with disabilities, women and girls in regions with limited digital access, ethnic and religious minorities, LGBTQ+ communities, and economically disadvantaged populations are rarely consulted, even though they are often most affected by AI-driven discrimination. How to include them: Targeted outreach and funding: Provide financial and logistical support to enable participation of underrepresented groups in AI Dialogue meetings. Regional pre-dialogue consultations: Organize preparatory meetings in different regions (Africa, Latin America, Asia-Pacific) to gather input before the main Dialogue sessions. Dedicated seats or quotas: Reserve specific speaking slots or representation quotas for civil society, broadly understood. Online and asynchronous participation tools: Use digital platforms that allow written contributions and virtual attendance, accommodating time zones and connectivity limitations. Partnerships with local organizations: Collaborate with regional networks (e.g., African Union, European Union, Latin American civil society platforms) to identify and amplify underrepresented voices.

The establishment of sustainable and continuous communication channels for relevant stakeholders at the international level should also be included as a separate heading, considering that it constitutes the basis for mutual understanding of AI-related matters. Moreover, the horizontal, dynamic nature and fast evolution of AI require ongoing communication and follow-up between all relevant actors. So dedicated online communication channels or groups (eg on Signal or other) with possibility of virtual meetings.

On top of online practices, the Dialogue should provide for a policy discussion forum for relevant stakeholders. Best practices of governments, different approaches of relevant stakeholders, civil society's in particular, on a specific matter should be discussed and reported in this forum without voting. The Dialogue should attach importance to close cooperation with the Independent Scientific Panel and thus aim to discuss and suggest evidence based policy preferences in the field of AI National and regional initiatives should be actively followed and common policy briefs should be prepared via cooperation with other international organizations. The Dialogue,should be the coordination hub for these activities due to its global reach.