StarApple AI | Adrian Dunkley | March 16, 2026

Small Island Developing States and the Global AI Governance Table: How Caribbean Nations Can Claim Their Seat

The global AI governance architecture is being built, conference by conference and treaty by treaty, and Caribbean nations are not yet builders.

International diplomacy and global governance representing Caribbean SIDS participation in AI policy

By March 2026, global AI governance has taken on a recognisable shape. The EU AI Act's high-risk AI obligations are approaching their August 2026 implementation deadline, with Caribbean governments and businesses affected by its extraterritorial provisions still working through their compliance positions. The Bletchley Declaration of November 2023, the Seoul AI Safety Summit of May 2024, the G7 Hiroshima AI Process, the OECD AI Principles, and the UN General Assembly's first AI resolution of March 2024: all of these formative moments in international AI governance occurred with minimal Caribbean SIDS participation. The frameworks that emerged from those moments will shape how AI is developed, deployed, regulated, and held accountable for decades.

My argument is plain. Caribbean SIDS must move from passive observers to active shapers of global AI governance frameworks, and that requires a deliberate shift in how Caribbean governments and regional institutions approach international engagement on AI. This is not an argument for Caribbean exceptionalism. It is an argument for Caribbean presence: at the tables where standards are set, at the forums where norms are negotiated, and in the institutions where implementation is designed.

This article examines why Caribbean SIDS have been marginalised in global AI governance, what specific interests the Caribbean needs to defend in international forums, and what a Caribbean strategy for global AI governance engagement looks like in practice.

The Marginalisation Pattern and Why It Is Not Inevitable

Caribbean SIDS' marginalisation in global AI governance is not primarily the result of deliberate exclusion. It reflects a pattern that recurs across global governance domains: technical forums begin as discussions among leading technology-producing nations, develop momentum and produce initial frameworks that reflect those nations' priorities, and by the time developing countries are formally included, the essential architecture is already fixed.

The Caribbean has faced this pattern in financial regulation, with the FATF's anti-money laundering framework. It has faced it in trade, with WTO agreements developed primarily by large economies. It has faced it in climate, where Caribbean SIDS were not at the table during the early formulation of the UNFCCC framework but have since, through the Alliance of Small Island States, become effective and recognised advocates for SIDS-specific concerns including the 1.5 degree Celsius temperature goal and loss and damage financing.

The climate comparison is instructive because it shows that late-entry SIDS advocacy can win real results when it is strategic, coordinated, and technically grounded. Caribbean nations' contributions to climate governance are disproportionate to their size. There is no structural reason the same cannot be true of AI governance. The difference between climate and AI is that the global AI governance architecture is less mature and less institutionally settled than the climate framework was by the time AOSIS became an effective force. The window for shaping AI governance fundamentals is still, just, open.

What Caribbean SIDS Actually Have at Stake

The case for Caribbean engagement in global AI governance is not abstract. There are concrete interests at stake that current international frameworks do not adequately protect.

The first is data sovereignty. Caribbean citizens' data is being processed by AI systems operated by companies headquartered in the United States, Europe, and elsewhere. The terms on which that processing occurs, the rights Caribbean citizens have to challenge AI-driven decisions made about them by foreign companies, and the liability regime that applies when those systems cause harm are all determined by frameworks in which Caribbean governments have had minimal input.

The second is climate-relevant AI. Caribbean SIDS are among the most climate-vulnerable territories on earth. AI systems for early warning, agricultural adaptation, fisheries management, and disaster response are potentially critical tools for Caribbean survival. The standards that govern how those AI systems are designed, tested, and held accountable need to reflect the specific requirements of small island climates and ecosystems.

The third is the economic dimension. Anguilla's .ai domain economy generated approximately US$39 million in 2024. The Caribbean's business process outsourcing sector, which serves North American and European clients, is increasingly AI-enabled. Tourism operators are adopting AI tools for demand forecasting, pricing, and customer engagement. Financial services in Trinidad and Tobago, Barbados, and Jamaica are AI-intensive. These are economic activities that depend on the international AI governance environment, and their practitioners have interests that deserve representation.

The fourth is the proportion of Caribbean AI adoption. At approximately 14 per cent, compared to 24 per cent in the Global North, the Caribbean's AI adoption rate reflects a technology access gap that global AI governance frameworks should address through measures on affordable access, capacity-building support, and non-discriminatory technology transfer.

International diplomacy documents representing Caribbean engagement in global AI governance

The UN System: Where Caribbean Voice Carries Most Weight

The UN system is structurally favourable to small states. One nation, one vote in the General Assembly. Universal membership. Formal processes for developing country participation. And a history of successful SIDS advocacy, from the Barbados Programme of Action (1994) to the SAMOA Pathway (2014) to the Bridgetown Initiative (2022) on multilateral development finance reform.

The UN General Assembly's first AI resolution, passed on 21 March 2024, demonstrates that the UN system can produce AI governance outcomes. The UN High-level Advisory Body on AI, whose final report, "Governing AI for Humanity," was published in September 2024, recommended the establishment of an international AI governance body within the UN system, with universal membership and attention to developing country needs.

Caribbean governments should be active supporters and shapers of whatever UN AI governance body emerges from these processes. I am arguing that CARICOM should seek to build an AI governance advocacy coalition with AOSIS members across the Pacific and Indian Oceans, creating a global SIDS AI governance coalition that no major power can ignore.

Engaging with the EU AI Act's International Dimensions

The EU AI Act contains provisions on international cooperation and is, through the EU AI Office, developing external engagement strategies. The Act's August 2026 deadline for high-risk AI obligations is approaching, creating both compliance pressures and diplomatic opportunities for Caribbean governments.

The EU-CARICOM relationship, grounded in the African, Caribbean and Pacific Group's historical partnership with the EU and the successor OACPS-EU Partnership Agreement, provides a diplomatic channel through which AI governance cooperation can be formalised. A structured EU-CARICOM AI cooperation agreement, providing for technical assistance, information-sharing on AI incidents, and Caribbean participation in EU AI standards development processes, would be a real win that serves both parties' interests.

Building Caribbean Technical Capacity for International Engagement

Effective international AI governance engagement requires technical credibility. A delegation that can only make general statements about developing country concerns will have less influence than one that can present technically precise proposals backed by evidence.

The University of the West Indies is the natural institutional home for Caribbean AI governance research and policy capacity. Its campuses in Jamaica, Trinidad and Tobago, Barbados, and across the OECS create a regional research infrastructure that could, with appropriate investment, generate the evidence base that Caribbean governments need to support credible international advocacy. A dedicated UWI Centre for AI Governance and Policy would serve this function.

Anguilla's .ai domain revenue is, in this context, a credible funding mechanism. A small percentage of .ai domain revenue, directed to a Caribbean AI Governance Capacity Fund, would generate a recurring, independent source of financing for Caribbean participation in international AI governance processes.

The AOSIS Model: Coalition Building for SIDS

The Alliance of Small Island States, established in 1990, transformed Caribbean climate governance advocacy from isolated national positions into a coordinated small island bloc that changed the architecture of the global climate regime. AOSIS was the force that put the 1.5 degree Celsius target on the agenda of the Paris Agreement negotiations. Its effectiveness comes from the combination of clear collective interests, disciplined coordination, and consistent technical engagement over decades.

I am advocating for the creation of an equivalent structure for AI governance: a Small Island Developing States AI Coalition that brings together Caribbean SIDS, Pacific SIDS, Indian Ocean SIDS, and the small states of the Atlantic and Mediterranean under a single advocacy framework. This coalition does not need to be a new institution from scratch. It can be built on the existing AOSIS structure by adding an AI governance working group. The CARICOM Secretariat should take the initiative in proposing this coalition and should volunteer to host its initial working group.

Recommendations

  1. Establish a CARICOM AI Ambassador with a mandate to represent the region in international AI governance forums. The Ambassador should be a senior official or eminent person with technical credibility in AI policy, mandated by CARICOM Heads of Government, and supported by a small secretariat within the CARICOM Secretariat.
  2. Propose a SIDS AI Governance Working Group within AOSIS at the next scheduled AOSIS meeting. CARICOM should bring a formal proposal to establish this working group with a defined mandate: to develop a common SIDS position on international AI governance and to coordinate SIDS participation in UN and other international AI forums.
  3. Initiate a formal EU-CARICOM AI Cooperation Dialogue through the OACPS-EU Partnership Agreement channel. The dialogue should seek structured consultation on the EU AI Act's implications for Caribbean SIDS, technical assistance for Caribbean AI governance capacity-building, and Caribbean participation in EU AI standards development processes.
  4. Establish a UWI Centre for AI Governance and Policy with regional mandate and international connectivity. The Centre should be funded through a combination of CARICOM member state contributions, international development assistance, and revenue from the .ai domain economy.
  5. Develop a Caribbean SIDS AI Governance Position Paper for submission to the UN AI governance process by the end of 2026. The position paper should articulate CARICOM's specific interests and advocacy priorities in international AI governance and be formally submitted to the relevant UN body.
  6. Explore a SIDS AI capacity-building fund proposal for the next G20 and OECD AI Policy Observatory engagement. The fund should provide financing for SIDS participation in international AI governance forums, SIDS government AI policy capacity-building, and SIDS AI research.

Conclusion

The global AI governance table has been set. The initial courses have been served: the Bletchley Declaration, the OECD AI Principles, the EU AI Act, the UN General Assembly resolution, the Seoul Safety Summit. Caribbean SIDS have observed these courses from outside the room. The question for 2026 and beyond is whether Caribbean governments will choose to take their seats at the table for the remaining courses.

The AOSIS model shows that late-entry SIDS advocacy can reshape an entire regime when it is strategic and sustained. The Caribbean has the intellectual capacity, the regional institutional infrastructure, and the concrete interests necessary to build an effective international AI governance advocacy programme. What it lacks is the political decision to treat this as a priority comparable to climate diplomacy and trade advocacy.

The concrete takeaway is this: appoint the CARICOM AI Ambassador, propose the SIDS AI Coalition, initiate the EU-CARICOM dialogue, fund the UWI Centre, and submit the Caribbean position paper to the UN. These are the steps that transform observer status into participant influence.

About the Author

Adrian Dunkley is a Caribbean AI governance expert with extensive experience in legal and regulatory framework analysis, legislative gap analysis, and policy reform recommendations in AI governance, digital technologies, data protection, and human rights law. He advises Caribbean government institutions and regional bodies on AI policy and has worked across Jamaica and the wider CARICOM region on digital economy development. Adrian is a co-founder of StarApple AI, the Caribbean's first AI company, and founder of AI Jamaica. Contact: insights@starapple.ai