From the Margins to the Centre: Caribbean Women as Leaders in AI Governance

The day after International Women's Day, it is tempting to write an article that celebrates what has been achieved. I am not going to do that, or not primarily. The gap between where Caribbean women are in AI governance and where the evidence says they should be is too wide to be closed by a celebration. It requires an argument.

My argument is this: Caribbean women are not only the population most likely to be harmed by unregulated AI, they are also the population best positioned to build more equitable AI governance, if the institutional barriers to their participation are deliberately dismantled. These are two sides of a single coin, and the connection between them is not coincidental. The communities most thoroughly excluded from the design and governance of AI systems are the communities whose exclusion is most evident in those systems' outputs. Including them in governance is therefore simultaneously an equity measure and a quality improvement.

The barriers are specific, and so are the fixes. I write as someone who has worked across the CARICOM region on these questions, who has sat in the governance rooms from which women's voices are too often absent, and who has watched what changes when those voices are present. What follows names the barriers and sets out what removing them would look like in practice.

The Argument from Expertise

Discussions of women's participation in AI governance often frame the issue as one of representation: women should be included because inclusion is fair. This framing is correct as far as it goes, but it is insufficient, and it tends to invite responses that treat women's inclusion as an accommodation rather than a substantive improvement. I want to make a different argument: Caribbean women's inclusion in AI governance is a technical quality requirement, not merely a fairness preference.

AI governance requires several types of expertise that Caribbean women disproportionately possess. The first is expertise in how AI systems fail marginalised communities. This expertise comes from direct experience: from being denied services by automated systems that were not designed with your circumstances in mind, from navigating digital platforms that assumed demographics different from your own, and from living with the downstream consequences of algorithmic decisions made without your input. This is not anecdote; it is evidence, and it is systematically underrepresented in the rooms where AI policy is made.

The second is expertise in the social systems AI is being asked to replicate or improve. AI systems for healthcare, social protection, education, agriculture, and financial services are being deployed in social contexts that Caribbean women understand with a specificity that most technologists and economists do not. A woman who has navigated Jamaica's public health system as a patient, a carer, and a community health advocate knows things about how that system actually functions, as distinct from how it was designed to function, that are essential for designing AI tools that will serve rather than harm.

The third is expertise in intersecting disadvantage. Caribbean women who face overlapping barriers of gender, race, class, and geography possess a sophisticated understanding of how multiple disadvantages interact and compound each other. This understanding is exactly what AI governance frameworks need to design systems that do not replicate those compounding disadvantages at scale. The WEF Global Gender Gap Report 2024 documents that these intersecting gaps persist across economic, educational, health, and political dimensions. AI governance that fails to understand how these dimensions interact will produce systems that fail at each intersection.

The Evidence of Harm That Makes This Urgent

The urgency of Caribbean women's AI governance leadership is directly proportional to the scale of AI's potential harm to those women in the absence of adequate governance. I have argued in previous work that AI hiring tools are encoding historical gender biases into automated employment decisions, that the digital gender divide will be converted by AI from a correctable inequality into a permanent structural feature, that gender data gaps in Caribbean statistics will produce AI systems that systematically underperform for women, and that existing legislative frameworks are inadequate to prevent or remedy these harms.

These arguments are not hypothetical projections. The EU AI Act's high-risk AI obligations are approaching full implementation in August 2026. The systems that will be classified as high-risk under Annex III, including AI in recruitment, healthcare, education, and social protection, are already deployed or in procurement across Caribbean markets. The compliance frameworks that would apply to their use in Europe do not apply to their use in the Caribbean. The women most affected by their failures have the least access to the governance processes that could prevent those failures.

The UN High-level Advisory Body on AI's final report, "Governing AI for Humanity", published in September 2024, explicitly calls for developing country representation in AI governance and for governance frameworks that centre equity. It does not call for women to be included as a separate consideration; it treats equity as the foundational organising principle of legitimate AI governance. This is the international consensus position. Caribbean governance frameworks should reflect it domestically.

What the Institutional Barriers Actually Look Like

Naming the barriers to Caribbean women's AI governance leadership is necessary before those barriers can be removed. In my experience across the CARICOM region, the barriers are consistent and specific.

The composition of technology advisory bodies and digital economy task forces in Caribbean governments skews heavily towards technologists, economists, and senior civil servants from line ministries with technology portfolios. Gender expertise, community advocacy expertise, and lived experience of marginalisation are systematically underrepresented. This is not usually a deliberate exclusion; it is the consequence of appointment processes that draw on professional networks in which women are underrepresented. The result is the same regardless of intent.

The timing and format of governance processes disadvantages women with care responsibilities. Advisory committee meetings scheduled at short notice during school hours, international AI governance meetings that require extended travel, and consultation processes conducted through written submissions that assume time and research support most women do not have, all create participation barriers that fall disproportionately on women. Flexible, funded, accessible participation mechanisms are not luxuries; they are prerequisites for representative governance.

The language and framing of AI governance discussions creates another barrier. Technical AI discourse, which dominates governance conversations, uses terminology and conceptual frameworks that presuppose familiarity with computer science and engineering. Women with legal, public health, education, community development, and social science expertise, all of which are directly relevant to AI governance, are treated as non-experts because they do not speak in the dialect of machine learning research. This is a category error that impoverishes governance.

The compensation structures for governance participation frequently do not reflect the opportunity cost for women, particularly women who are early-career researchers, community advocates, or consultants without institutional backing. Unpaid advisory roles, nominal per diems that do not cover genuine participation costs, and governance processes that demand heavy preparatory work without adequate support all create financial barriers that fall hardest on women.

The Caribbean Advantage: Small Scale, High Potential

There is a structural advantage to Caribbean AI governance that is almost never discussed: the region's small scale is a genuine asset for participatory, representative governance. In Jamaica, a country of approximately three million people, the policy community is small enough that government ministers genuinely know the country's leading researchers, civil society leaders, and practitioners across multiple domains. A determined policy initiative can achieve genuine stakeholder inclusion in a timeframe and at a cost that would be impossible in a country of fifty million.

This means that the Caribbean's democratic deficit in AI governance, the gap between who is affected by AI and who participates in governing it, is not an intractable structural problem. It is a political choice. The choice to convene technology advisory bodies that do not include gender experts, legal scholars, public health specialists, and community advocates is a choice. Reversing it is a choice of equivalent simplicity, though it requires political will that the region's AI governance bodies have not yet consistently demonstrated.

The University of the West Indies, with campuses in Jamaica, Trinidad and Tobago, Barbados, and affiliated centres across the CARICOM region, represents another structural asset. UWI has the intellectual resources to build genuinely interdisciplinary AI governance capacity that combines law, gender studies, data science, public health, and public policy. Several UWI faculties have begun work in AI ethics and digital governance. The missing element is the institutional decision to make this work a regional priority, funded at a level commensurate with its importance.

The OECS Commission and CARICOM Secretariat provide regional coordination mechanisms that can translate national governance improvements into regional standards more efficiently than in most other regions of the world. When Jamaica develops a gender-responsive AI procurement standard, it can be adapted for Barbados, Trinidad, and the OECS members within months rather than years, because the regional institutional infrastructure exists. The Caribbean is, in governance terms, a place where good ideas can travel fast.

The International Moment and Caribbean Positioning

The international AI governance moment is unusually favourable for Caribbean leadership. Gender and AI is emerging as a distinct policy area globally, with increasing attention from UN Women, UNESCO, the ITU, and the academic community. Caribbean scholars and policymakers who position themselves as contributors to this emerging field now will have a disproportionate influence on the direction that field takes.

The EU AI Act's approaching high-risk obligations, due for full implementation in August 2026, create a specific opportunity. European AI vendors seeking Caribbean market access will face increasing pressure to demonstrate that their equity compliance extends beyond European diversity standards to include the populations of countries where their systems are deployed. A Caribbean AI governance framework that specifies equity requirements for market access gives Caribbean governments real bargaining power in procurement negotiations, power that does not currently exist because the framework does not yet exist.

The CEDAW framework, to which all CARICOM states are parties, provides an international accountability mechanism that the region barely uses. CEDAW's interpretation has grown to cover women's participation in the governance of public technologies. Caribbean states that report on their AI governance practices in CEDAW periodic reviews, and that face questions from CEDAW committees about women's representation in AI governance bodies, gain another reason to show real progress. Using international human rights mechanisms to press for domestic AI governance reform is a strategy that Caribbean women's rights advocates have not yet pursued consistently, but it is available.

Recommendations

1. Establish mandatory gender parity requirements for all government AI advisory bodies. Every national AI advisory committee, digital economy task force, or technology governance body in CARICOM member states should be required to achieve gender parity in membership as a condition of formal establishment. Gender parity should be defined to include women with diverse professional backgrounds, not only women from technology fields. The CARICOM Secretariat should develop a model governance code for regional and national AI bodies that includes this requirement.
2. Fund a Caribbean Women in AI Governance Fellowship programme. A regional fellowship programme, administered jointly by the University of the West Indies and a regional body such as the OECS Commission, should place women with relevant expertise in AI governance roles in Caribbean governments and regional institutions for 12-month periods. Fellows should combine their governance placement with research into gender and AI policy, producing work that feeds directly into regional policy development. Funding should be sought from the Caribbean Development Bank and international development partners.
3. Reform governance participation structures to remove practical barriers. All Caribbean government AI governance bodies should adopt participation standards that include: flexible meeting schedules and remote participation options; full reimbursement of participation costs including childcare; plain-language governance documentation accessible to non-technical participants; dedicated support for participants who need to develop technical literacy to contribute effectively; and explicit recruitment processes targeting women from legal, health, education, community development, and social science backgrounds.
4. Commission a Caribbean women's AI leadership audit from the University of the West Indies. UWI, in partnership with UN Women Caribbean, should conduct a systematic audit of women's representation in AI governance roles across CARICOM, including government advisory bodies, regulatory authority leadership, technology industry boards, and academic AI research groups. The audit should identify specific institutional barriers and produce tailored recommendations for each type of governance body. Results should be published and presented to CARICOM Heads of Government.
5. Use CEDAW periodic reviews as an AI governance accountability mechanism. Caribbean states due for CEDAW periodic reviews should be required, as a matter of government policy, to include a substantive section on women's participation in AI governance and on AI-related gender discrimination risks. Civil society organisations, including women's rights NGOs and technology advocacy groups, should be supported to submit shadow reports on AI governance gender equity for consideration by the CEDAW Committee.
6. Create a Caribbean Women's AI Governance Network with institutional support. A formal network connecting women working on AI governance across the CARICOM region, supported by the CARICOM Secretariat and hosted by UWI, would build the peer connections, shared knowledge, and collective advocacy capacity that individual women in governance roles currently lack. The network should connect to international bodies including the Global Partnership on AI, the ITU Focus Group on AI, and UN Women's technology programme, positioning Caribbean women's voices in global AI governance conversations.

A Design Principle, Not a Courtesy

The argument here is not simply that Caribbean women deserve a seat at the AI governance table, though they do. It is that AI governance without Caribbean women's expertise, perspective, and leadership will produce inferior outcomes, not only for Caribbean women but for Caribbean societies as a whole. Systems designed without the input of the communities they affect most will serve those communities least effectively. This is not a controversial claim; it is a design principle.

The institutional barriers that keep Caribbean women at the margins of AI governance are not immovable. They are specific, named, and addressable by deliberate policy action: gender parity requirements for advisory bodies, funded participation mechanisms, reformed appointment processes, accountability through CEDAW, and investment in UWI's capacity to build interdisciplinary governance expertise. None of these is an ambitious vision. They are actions Caribbean governments can take this year.

The AI governance frameworks that Caribbean societies build in the next three years will shape how AI affects the region for the next thirty. Building those frameworks without Caribbean women at the centre, both as the population whose experiences must be protected and as the leaders whose expertise must shape the governance, is not a neutral technical choice. It is a political decision with predictable consequences, and those consequences will be paid most dearly by the women who were shut out of making it. So the test for any minister reading this is simple: look at who is in the room the next time an AI advisory body is convened, and ask who is missing and why.

About the Author

Dr. Shirley Budall is a Caribbean expert in gender, inclusion, and AI governance with demonstrated experience in the ethical, legal, social and governance dimensions of artificial intelligence and digital technologies. She conducts legal and regulatory framework reviews and develops policy recommendations for legal reform in AI governance, data protection, human rights, and gender equality. Dr. Budall has knowledge of international and regional AI governance standards and has advised Caribbean government institutions and regional organisations on inclusive AI policy. She is a researcher and consultant working across the CARICOM region on digital economy governance, women's rights in the digital age, and equitable technology development. Contact: insights@starapple.ai