n/a

From my perspective, a successful first Global Dialogue on AI Governance would deliver outcomes that are practical, inclusive, and directly applicable to real-world implementation. Drawing on my experience in cybersecurity operations, compliance, and outreach, success depends on moving beyond high-level discussion to clear, actionable results . First, the Dialogue should produce agreed baseline standards for safe, secure and trustworthy AI, particularly for high-risk systems. In cybersecurity, clearly defined controls and procedures are essential for managing risk, and similar principles should guide AI governance. Second, success would include established mechanisms for international cooperation, especially for incident reporting, threat intelligence sharing, and coordinated response. AI-related risks, like cyber threats, are cross-border, and effective governance requires structured collaboration. Third, inclusive participation and capacity-building should be a core outcome. Through my work in training and community programmes, I have seen how strengthening skills and awareness improves resilience. Supporting developing countries and under-resourced organisations will reduce global vulnerabilities and ensure more equitable adoption of AI. Fourth, the Dialogue should bridge the gap between policy and practice by actively involving technical practitioners. My experience developing playbooks and supporting operational security highlights the importance of governance that reflects how systems are actually monitored, secured, and maintained. Finally, success would be measured by sustained momentum, including clear next steps, timelines, and ongoing collaboration platforms. This would ensure the Dialogue becomes a foundation for long-term international coordination rather than a one-time event.

• Safe, secure and trustworthy AI
• Transparency, accountability, and human oversight
• Interoperability of governance approaches
• AI capacity-building

From my perspective, these four areas reflect the most urgent priorities for practical and effective AI governance, based on my experience in cybersecurity operations, compliance, and training . Safe, secure and trustworthy AI is foundational. AI systems introduce new security risks such as data poisoning, model manipulation, and automated misuse. Addressing these risks requires embedding security controls across the entire lifecycle, similar to established cybersecurity practices. Transparency, accountability, and human oversight are critical for managing and responding to incidents. In practice, effective security depends on visibility into system behaviour, clear audit trails, and defined responsibility. These elements support faster detection, investigation, and remediation when issues arise. Interoperability of governance approaches is essential because AI risks are inherently cross-border. Fragmented regulatory frameworks create gaps that can be exploited and make coordinated responses more difficult. Aligning standards and approaches improves consistency, reduces uncertainty, and strengthens collective resilience. AI capacity-building is equally important. Through my work in outreach and training, I have seen how gaps in skills and resources directly impact security outcomes. Supporting developing countries and organisations ensures more consistent implementation of safeguards and reduces systemic vulnerabilities.

Yes, there are several cross-cutting and emerging issues that are not fully captured by the listed themes and are important from a cybersecurity perspective. One key gap is AI supply chain security. AI systems rely on complex ecosystems that include datasets, pre-trained models, third-party tools, and cloud infrastructure. Each component introduces risk, and compromise at any stage can affect downstream systems. Governance efforts should address integrity, provenance, and third-party risk across the full lifecycle. Another issue is adversarial misuse of AI. While safety is referenced broadly, there is limited focus on how AI is actively being used to scale cyber threats such as phishing, social engineering, and automated exploitation. This requires stronger coordination between security communities and AI governance processes. Incident response for AI systems is also underdeveloped. In cybersecurity, structured response processes are critical, yet AI governance discussions often focus more on prevention than on how to handle failures or attacks. Clear protocols for detection, reporting, containment, and recovery are needed. There is also a growing need for standardised evaluation and assurance methods. Organisations require consistent ways to test AI systems for robustness, security, and reliability both before and after deployment. Finally, concentration of compute and infrastructure presents a systemic risk. A small number of providers control much of the infrastructure used to develop advanced AI, which creates potential single points of failure and raises questions about resilience and equitable access. These issues cut across existing themes and highlight the need for a more operational and security-focused approach to AI governance.

From my perspective, governance gaps in safe, secure and trustworthy AI, transparency and accountability, interoperability, and capacity-building are already shaping both challenges and opportunities within my sector and region. A key challenge is the rapid adoption of AI without consistent security standards. In practice, organisations are integrating AI into operations faster than governance frameworks are evolving, which increases exposure to risks such as data leakage, model manipulation, and automated social engineering. Limited transparency and unclear accountability further complicate incident investigation and response, making it harder to trace system behaviour or assign responsibility when issues arise. Fragmentation across governance approaches also presents a challenge. Differences in regulatory expectations between countries create uncertainty and can leave gaps that threat actors exploit. For regions like the Caribbean, this can slow coordinated responses and make it harder to align with global standards while addressing local needs. Capacity constraints remain a significant issue. Many organisations face limitations in technical expertise, infrastructure, and access to specialised tools. From my experience in training and outreach, these gaps directly impact the ability to implement effective security practices and maintain resilience across systems . At the same time, these challenges create important opportunities. There is growing momentum to embed security and accountability into AI systems from the start, drawing on established cybersecurity practices. Increased attention to governance is also encouraging stronger collaboration between public and private sectors, as well as investment in skills development and awareness. Overall, these developments highlight the need for practical, coordinated approaches that strengthen resilience while supporting responsible adoption of AI.

From my perspective, the AI Dialogue can play a key role in advancing international cooperation by creating practical pathways for coordination, not only discussion. Drawing on my experience in cybersecurity operations, compliance, and outreach, effective cooperation depends on shared structures, clear communication, and consistent implementation across different environments . One important role is establishing common approaches to incident response for AI systems. Similar to cybersecurity, countries need aligned processes for identifying, reporting, and responding to AI-related risks. This would improve coordination and reduce response times when incidents occur across borders. The Dialogue can also strengthen cooperation through structured information sharing. In practice, detecting and understanding threats relies on access to diverse data sources. Expanding this to include AI vulnerabilities, misuse patterns, and lessons learned would help countries anticipate risks earlier and respond more effectively. Another key contribution is supporting capacity-building and knowledge exchange. Through my experience in training and community engagement, I have seen how improving technical skills and awareness leads to stronger security outcomes. The Dialogue can help ensure that developing countries and smaller organisations are equipped to implement governance measures and participate fully in global efforts. Finally, the Dialogue can help align policy with operational reality by including technical practitioners in governance discussions. This ensures that frameworks are realistic, implementable, and informed by how systems are actually secured and managed. If effectively structured, the AI Dialogue can move beyond general cooperation and become a platform for coordinated, sustained action on AI governance.

From my perspective, the AI Dialogue should build on existing cybersecurity and AI governance initiatives that already support collaboration, standard-setting, and capacity-building, while adding stronger coordination and practical alignment. One key area is established cybersecurity frameworks such as ISO 27001 and the NIST Cybersecurity Framework. These provide structured approaches to risk management, incident response, and compliance that can be adapted to AI systems. In my experience working with these frameworks, their strength lies in being actionable and widely recognised . The Dialogue should also connect with international cooperation mechanisms such as national Computer Security Incident Response Teams and global threat intelligence-sharing networks. These systems already enable cross-border coordination on cyber incidents and can be extended to include AI-related risks, vulnerabilities, and misuse patterns. In addition, partnerships with organisations focused on AI governance, research, and standards development are important. These include multi-stakeholder initiatives that bring together governments, industry, and civil society to address ethical, technical, and policy challenges. The added value of the AI Dialogue lies in its ability to bring these efforts together into a more cohesive global approach. Currently, many initiatives operate in parallel, which can lead to fragmentation and duplication. The Dialogue can help align standards, promote interoperability, and ensure that governance approaches are consistent across regions. It can also strengthen inclusion by connecting global frameworks with regional and local efforts, particularly in developing countries, and by ensuring that practical implementation challenges are reflected in global discussions. Overall, the Dialogue can act as a bridge between existing initiatives, turning fragmented efforts into coordinated, scalable action.

Different stakeholders can contribute to the AI Dialogue by bringing both policy insight and practical experience, with a structure that supports collaboration and implementation. Governments should provide policy direction, align national strategies, and support international coordination. Industry can contribute technical expertise, real-world use cases, and insights on system design, deployment, and risk management. Civil society and academia play an important role in raising ethical considerations, conducting independent research, and ensuring accountability. Technical practitioners, including those in cybersecurity, should contribute operational knowledge on how systems are monitored, secured, and maintained in practice, based on hands-on experience developing procedures, analysing threats, and supporting response efforts . To be effective, the Dialogue should adopt a structured, multi-layered format. High-level plenary sessions can set priorities and direction, while smaller working groups focus on specific areas such as security, governance standards, and capacity-building. These groups should include a mix of policymakers and practitioners to ensure outcomes are both strategic and implementable. The Dialogue should also include regular knowledge-sharing sessions, where stakeholders present case studies, lessons learned, and emerging risks. This would encourage transparency and continuous learning. In addition, there should be clear mechanisms for follow-up. This could include defined timelines, progress tracking, and periodic reviews to ensure that commitments lead to measurable outcomes. Finally, inclusive participation is essential. The structure should support engagement from developing countries, youth, and underrepresented groups to ensure diverse perspectives are reflected. If designed this way, the AI Dialogue can move from discussion to coordinated, sustained action.

Black young women remain significantly underrepresented in global discussions on AI governance, despite being directly affected by the social, economic, and technological impacts of AI. This gap limits the diversity of perspectives shaping policies and risks reinforcing existing inequalities in how AI systems are designed and governed. From my perspective, inclusion starts with recognising that representation is not only about participation but also about influence. Black young women bring valuable insight at the intersection of technology, community impact, and lived experience. In my own work across cybersecurity, AI engagement, and community programmes, I have seen how access to training, mentorship, and leadership opportunities can shift both confidence and outcomes . To address this gap, targeted capacity-building programmes are essential. Initiatives that provide technical training, scholarships, and exposure to AI and cybersecurity careers can help build a stronger pipeline. Mentorship and sponsorship programmes are also important to support progression into leadership and decision-making roles. Inclusion should also be built into the structure of global dialogues. This means actively inviting and funding participation from young professionals, especially from underrepresented regions, and ensuring their contributions are reflected in outcomes, not treated as symbolic. Community-based organisations and youth-led networks should be engaged as partners, not only participants, to ensure discussions are grounded in real-world impact. Creating safe and accessible spaces for contribution, both online and in person, will also improve engagement. By intentionally including Black young women, AI governance becomes more equitable, informed, and effective.

To foster meaningful and dynamic engagement, the AI Dialogue should centre real human experiences and create space for people to share how AI is affecting their lives and work. One effective format is lived-experience storytelling sessions, where individuals from different regions and sectors share how AI systems impact them directly. This could include young professionals, community leaders, and practitioners explaining challenges such as bias, access gaps, or security risks. Grounding discussions in real experiences helps move conversations from theory to reality. Small, facilitated dialogue circles can also support deeper engagement. In these settings, participants from different backgrounds speak openly about local challenges and priorities, while facilitators ensure that all voices are heard and captured. This format works well for including underrepresented groups and building mutual understanding. Another approach is interactive problem-solving sessions based on real cases. Participants can work through actual scenarios, such as responding to AI misuse or addressing gaps in access and skills. From experience in cybersecurity and training, practical exercises like these lead to clearer, more usable outcomes. Youth and community-led sessions are also important. Allowing young people, especially from underrepresented groups, to design and lead parts of the Dialogue ensures their perspectives are not filtered or overlooked. Finally, feedback loops should be built in. Participants should see how their contributions shape outcomes, which builds trust and encourages continued engagement. Focusing on real human experiences makes the Dialogue more inclusive, grounded, and impactful.

Several existing policies and practices already offer practical approaches to effective AI governance, especially when viewed through a cybersecurity lens. Frameworks such as ISO 27001 and the NIST Cybersecurity Framework provide strong foundations for managing risk, ensuring accountability, and maintaining continuous monitoring. In my experience working with these standards, their value comes from clear processes for identifying, assessing, and mitigating risks, which can be adapted to AI systems to strengthen security and reliability . Risk-based regulatory approaches are also effective. For example, classifying AI systems based on their level of impact allows organisations to apply stricter controls to high-risk use cases. This mirrors established cybersecurity practices where critical systems require stronger protections and oversight. Another important practice is the use of red-teaming and independent testing. Actively testing AI systems for vulnerabilities, misuse, and unintended behaviour before and after deployment helps identify weaknesses early and improves resilience over time. Transparency mechanisms such as audit logs, documentation standards, and impact assessments also support governance. These practices make it easier to understand how systems operate, investigate incidents, and ensure accountability. On the collaboration side, threat intelligence sharing platforms used in cybersecurity offer a strong model. Expanding these to include AI-related risks would improve collective awareness and enable faster, coordinated responses. Finally, capacity-building initiatives, including training programmes and community outreach, are essential. From my experience supporting training and awareness efforts, building skills and understanding at all levels strengthens implementation and long-term resilience . Together, these approaches show that effective AI governance is achievable when grounded in practical, tested methods.