Embassy of the Republic of Indonesia

A successful transition will involve a shift away from technical governance towards a sociotechnical & relational framework. AI governance has traditionally been measured in terms of algorithmic transparency, but for post-conflicted societies such as Lebanon, the greatest threat to societal stability is not only based on the ability of algorithms to create bad code; rather, it is the neurobiological effects that will result from the AI narratives produced by their use. Success should be determined by: - Including measures of "relational safety": this should go beyond the conventional measures of data privacy to assess how the governance of AI can assist to mitigate the Infocratic polarization of fragile states. - Committing to Digital Democratic Infrastructure (DDI): This requires the establishment of international standards for participatory AI that facilitates the ability of local communities to devise their own tools for Digital Peacebuilding. - Cross-Regional Knowledge Transfer: Bridging the gap between digital-first nations (such as Estonia) and societies that have experienced historical trauma (such as Lebanon) will assist in ensuring that AI governance is contextually and emotionally sensitive. The Dialogue will succeed to the degree that it acknowledges that AI governance is at the same time a regulatory challenge, and a basic necessity for building narrative resilience and achieving the Sustainable Development Goal of SDG 16 (Peace, Justice and Strong Institutions).

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

My selection is informed by my research into Intercultural Emotional Intelligence (ICEI) and its application to digital diplomacy. - Social & Cultural Considerations: AI it has a powerful impact on the "lived experience" of societies. In post-conflict areas AI generated imagery may cause a collective "threat response" (the Amygdala) which traditional governance does not take into account. There is a need to put linguistic and cultural considerations for AI at the forefront of discussions to avoid "digital sectarianism". - Capacity Building: Nations in the Global South must not just be consumers of AI Governance. The focus of any Capacity Building should be on supporting local "Artivists" and researchers in developing Digital Democratic Infrastructure reflective of their unique Social Context. - Human Rights & Human Oversight: My "Neuro-Diplomacy" concept supports viewing "Human Oversight" through a broader lens than strict technical audits, "Intercultural Emotional Intelligence" must also be included in determining if AI Systems are in anyway violating the dignity of marginalised individuals or furthering historical sensitivities. Transparency: Accountability requires accountability to include transparency within the algorithm's "black box", and all intents and effects of the narrative disseminated. Governance will ensure that AI supports and promotes relational engagement rather than transactional polarities.

Yes: The Neuro-Biological Impact of AI-Driven Information Ecosystems. The present emphasis of governing AI has primarily been upon the "safety" and ethics" of AI from a legalistic standpoint. At this time, however, there is an emerging crisis of "cognitive and emotional security" through the increasing engagement-oriented optimization of AI models, with those models serving to amplify emotional stimulus in weak state conditions, thus circumventing the prefrontal cortex and triggering the survival circuitry of the brain. I recommend that Neuro-Diplomacy be an additional theme across all AI Governance issues which would include: - The measurement of "emotional toxicity" of AI produced outputs in sensitive cultural settings - The development of "relational benchmarking" for AI, whether an A.I. aids in generating empathy or whether the use of the AI creates hostile "out group" response to AI use - Developing institutional resilience through providing training on ICEI for diplomatic and civic Leaders to support their ability to manage the sociotechnical interface between humans and A.I. By addressing the Neuro-social dimension of AI Governance the UN will transition from protecting "Data" to protecting the Human Social Fabric.

In Lebanon, and in the greater Mediterranean region, the main governance gap is a lack of socio-institutional safeguards that take global ethical principles and convert them into the operational realities of fragile states. While there is a debate in Global Northern countries about issues related to transparency surrounding algorithms, our part of the world has "accountability asymmetry" because we are consumers of AI-enabled platforms that have been designed for very different cultural understandings than the ones found within our region, and that inadvertently amplify sectarian "emotional triggers" and historical traumas. Without governance at the local level, these systems have increased the level of narrative warfare, and therefore put at risk the already fragile social fabric of most post-civil war countries. Challenges we face as a result of Information Poisoning and Cognitive Security - Narrative Dysregulation: The lack of "human oversight" metrics that are finely tuned to be Intercultural Emotionally Intelligent (ICEI) means that AI is being optimized for engagement without regard to the "triggering" of a collective threat response (Amygdala) that results from polarizing "slop" and/or disinformation. - Infrastructure Gaps: The urban/rural digital divide inhibits equitable access to AI in education and smart agriculture, thereby reinforcing existing social inequities. The Gaps Provide Opportunities for Lebanon The gaps in resilience have presented Lebanon with a unique opportunity to transform itself from a consumer to an architect of narrative resilience. The Emergence of Sovereignty Through AI Development The development of Digital Democratic Infrastructure (DDI) provides an opportunity for Lebanon to create a participatory governing model whereby civic leaders and activists collaborate in the design and development of artificial intelligent systems in such a way that Relational Engagement is prioritized over transactional metrics. Leadership Through Neuro-Diplomacy By embedding my research on ICEI into the theoretical foundations of institutional building, we will create a new model of governance that will utilize artificial intelligence to predict and mitigate sources of social friction; thereby transforming artificial intelligence from a tool of polarization to a means of promoting social cohesion and SDG 16. The challenge is that the current form of governance is "global in nature, but still rooted in the geopolitical realities of each region." The opportunity is to leverage the bottom-up approach to governance by providing for the validation of the "lived mark" of the citizens of Lebanon; and, thus, ensuring that AI will support our democratic future rather than undermine it.

The potential of AI Dialogue as a link between the disjointed "islands" of AI Governance worldwide is a unique one. While Silicon Valley creates new technical standards for the AI industry and Brussels creates new regulatory frameworks for AI governance, the "socio-tech" impact is being distributed unevenly; however this reality is particularly acute in "fragile states". Therefore, the AI Dialogue has great potential to enhance international collaboration by expanding beyond a focus only on technical interoperability and considering ways to establish Institutional and Narrative Interoperability among the various nations of AI Governance. The AI Dialogue can be pivotal in achieving the following: - Harmonizing "safe AI": The AI Dialogue can convene to develop a plan to achieve a globally accepted definition and standard of what constitutes "safety" for AI through the creation and adoption of Intercultural Emotional Intelligence (ICEI) to define "safety" consistently across all cultures, languages, and settings. This will enable the UN to promote the implementation of "Safe AI" technologies in various global settings, including, for example, that "Safe AI" refers to the same or very similar level of "Safety" in Estonia as it does in Lebanon. - Mitigating the "Digital Sovereignty" divide: The AI Dialogue can provide a neutral forum for nations to engage in collaborative dialogue and communication to prevent the development of a "Digital Cold War" between nations, and that through the AI Dialogue, nations will develop their AI capacity based on collaborative/nation-to-nation transfer of Digital Democratic Infrastructure (DDI) blueprints rather than through purely transaction-based transfers. - Validating and Supporting "South-to-North" Perspective: By bringing together researchers from post-conflict areas who have been subjected to disinformation, sectarianism, and cognitive warfare, the same global challenges now emerging, the AI Dialogue can assist in creating "Stress-test" methodology by which global governance can be validated and evaluated to ensure they have endured such conditions, in order for those global governance entities to serve as way to ultimately develop and implement rules, regulations, standards, and processes to govern the use of AI in a globally cooperative and coordinated manner. The Dialogue's main goal is to create a "Global AI Operating System" that has empathy and resiliency. In addition, it wants to convert AI Governance from being a point of geopolitical rivalry into a common infrastructure for SDG 16.

AI Dialogue is intended to serve as the "connective tissue" linking various established frameworks. AI Dialogue should be aligned with: - UN Global Digital Compact - To establish AI Governance connected to Universal Human Rights and Digital Inclusion; - UNESCO Recommendation on the Ethics of AI - By using the framework's focus on cultural diversity and linguistic rights as tools to achieve AI Ethics; - Institute for Economics & Peace (IEP) Frameworks - By implementing "Positive Peace" metrics in order to identify and measure the socio-technical effects of AI as it relates to State Resilience; - Regional Hubs including the ActEU - By using European-based initiatives focused on "Act for Democracy" that will enable the scaling of digital tools for democracy globally. The Added Value of the AI Dialogue The AI Dialogue is distinguished by its Universal Legitimacy/Collective Governance as a unique Value Proposition. While institutions such as the OECD or G7 are available for Technical Guidance, the options provided are often not representative of the Lived Experiences of participants from the Global South. The AI Dialogue provides: - A Multi-Stakeholder "Backstage" - A safe place where Governments, "Artivists" and Sociotechnical Researchers can create new relationships and collaborate without traditional power hierarchies. - The "Neuro-Diplomacy" Perspective - AI Dialogue introduces the notion of Intercultural Emotional Intelligence (ICEI) into UN policy language to fill the Cognitive Security gap created by many existing Technical Frameworks. - Institutional Continuity: The 2026/2027 roadmap offers a blueprint for Policymaking Process, as opposed to the "one-off summits," for example, the Digital Democratization Infrastructure (DDI) was proposed as a global pilot instead of just something that existed only on paper.

In order to achieve true inclusiveness in AI Dialogue, it is important to develop a more comprehensive model beyond the current "Plenary-only" approach. As an alternative, I suggest a modular system with three distinct but interrelated types of engagement: - The Research Layer: An International AI Observatory will enable researchers, particularly from the Global South, to submit stress tests of AI models in culturally sensitive contexts. - The Civic Layer: The Civic Layer will have a citizens' assembly-like structure, where stakeholders such as SDSN Youth can participate in discussions about the long-term effects of AI on work and civic engagement across generations. - The Governmental or Diplomatic Layer will engage in "Peer to Peer Learning" instead of having nations dictate to each other what constitutes "best practices." Mid-sized nations, or "Digital Mid-Powers" should lead workgroups focused on finding interoperable governance models. Format Suggestions: - Hybrid Policy Sprints - Short bursts of collaborative working sessions, typically involving technologists plus social science researchers, to work together and identify the cultural biases inherent in various AI models. - Narrative Submissions - Provide a method for individuals & organizations outside of traditional academic methods to share their experiences/observations related to the lived experience of their culture by allowing for the sharing of creative/artist-reflected narratives or examples of local digital democratic infrastructures. - Regional Satellite Dialogues - Host multiple energy-building events for groups in Arabia (i.e. Beirut), Africa (i.e. Nairobi), and Central America (i.e. San Salvador) to help ensure that the Geneva outcomes accurately reflect the regional context of the participants.

Right now, global conversations about artificial intelligence (AI) are largely based on two sets of opinions: those of the "AI Superpowers" (with an emphasis on technical/economic aspects) and those of EU regulators (with an emphasis on regulations). The groups that often don't have a voice at these discussions are members of the post-conflict and fragile countries where the socioeconomic effects of AI are immediate, physical, and cognitive threats to their safety and survival. The following groups are some of the most underrepresented in AI discussions: 1. Non-Western Language Communities: AI governance typically lacks consideration for the differences in communication/usage between "small-languages" or dialects (for example, Levantine Arabic), which negatively impacts moderation and creates an increase in "digital sectarianism", etc. 2. "Artivists" From the Global South: People who are using evolving creative technology to support peacebuilding are frequently excluded from incidents of high-level governance. 3. Young People Living In Conflict Zones: The demographic group that is most severely affected by AI-driven polarization is generally the least consulted or represented in discussions about AI. How to include them: - Travel & Participation Grants: Add explicit funds for independent researchers and civil society from Global South that will pay their way to participate. - Linguistic Diversity Mandates: Produce all Dialogue output/input in more than just the 6 UN languages using AI technology to bridge the language gap. - Lived Imprint Registry - Develop a formal UN mechanism that allows communities to report 'algorithmic harms' and/or 'cultural bias' on a direct pipeline to Co-Chairs.

Many of the traditional United Nations panels are merely transactional, as the presenters come to discuss their respective "pre-packed" positions, leaving little room for a synthesis of the individual discussions. To create more meaningful participatory engagement, we must create relational formats: - Neuro-Diplomacy Simulations: Role-playing simulations in which participants are responsible for managing an AI-driven crisis in a fictitious post- conflict country can help attendees develop and quickly implement Intercultural Emotional Intelligence (ICEI) throughout the experience. - Artivism Digital Galleries: Using technology, immersive digital galleries should be created that could accompany the Geneva meetings and offer illustration/visualization of the "invisible" effects of AI such as data labor in the Global South or "Echo Chamber" architecture, ultimately helping to establish connections between the emotions of the participants and the reality of the advocacy being discussed. - Algorithmic Town Halls: These tools will provide representatives of the global public the opportunity to vote on their policy priorities using participatory AI tools such as Pol.is, prior to and during the sessions. These votes will create "Real-Time Feedback Loops" between the Geneva delegates and the global "Digital Democratic Infrastructure," providing for a more engaged public. - Backstage Roundtables: These are Chatham House-style sessions in which CEOs and grassroots activists can meet behind closed doors to discuss the "Human Oversight" gap. The use of the formats above should shift the dialogue from a "Talking Shop" to a "Living Laboratory" for producing global social cohesion.

To fill the governance voids in states where there is instability, we need to reference models that value Human Agency and Relational Security. Below are four tangible examples of how this can be accomplished: - Digital Democratic Infrastructure (DDI): A participatory model that sees the digital environment as a public resource. Rather than create new regulations from the top down, DDI leverages AI as a platform for establishing "Civic Sprints" or a digital version of a town hall, using clustering algorithms similar to Pol.is to find consensus around differences in religious and ethnic affiliations. This changes the use of AI from a polarising and divisive device into an engaging device for creating collective agreement about everyday issues. - The "Intercultural Emotional Intelligence (ICEI) Metric": This process takes the combined uses of Emotional Intelligence and Cultural Intelligence and incorporates them into the "Red-Teaming" process of AI development. Before any model is deployed into a sensitive area, such as the Levant, it should be "stressed" against its "Lived Imprint" to ensure that the moderation algorithms do not cause, through inadvertent means, the triggering of historical trauma or religious sectarian responses. - The "Neuro-Diplomacy" Protocol: This practice is to develop AI Auditors who are knowledgeable about the neuroscience of threat behaviour as a mechanism for auditing algorithms for "Emotional Toxicity" in addition to the traditional method of auditing algorithms for "Toxic Keywords." This will enable us to identify systems that favour high arousal (and thereby high polarization) content and help reduce social conflict resulting from these types of systems. A Gold Standard Model of Interoperable and Secure Data Exchange UMS + Collective Action: Estonia's model of interoperable, secure data exchange is a worldwide leader in this category. However, utilizing Artivism in conjunction with the X-Road in post-conflict regions, provides citizens the opportunity to use creative technological platforms to visualise and process trauma in more generalised or traditional healing practises. The shift from a purely algorithmic focus to a more human-centred interface-based model for building ICEI and DDI will create the groundwork for developing a truly "Resilient AI" approach that not only safeguards data but also strengthens the social cohesion of countries and regions.