This is my personal post for International Women's Day 2026, written as a Jamaican man who has worked in AI for more than fifteen years and has watched Caribbean women work in and around this field. It is not an institutional statement.
My choice today is to name and back the Caribbean women building AI careers, companies and projects without waiting for an invitation. Many of them walked into rooms that were not set up for them and built things anyway.
Where I have seen women underestimated
Since I founded StarApple AI in 2016, the women who have engaged seriously with AI in our programmes have often been among the strongest practitioners, and they have been underestimated at each stage:
- Applying for training, where some still assume AI suits men better.
- Proposing projects, where their technical credibility is questioned in ways male colleagues do not face.
- Delivering results, where credit is not always shared fairly.
- Speaking in public, where some audiences still doubt that a Jamaican, Trinidadian or Barbadian woman can be a serious AI professional.
The women who got past this did it through excellent work, persistence, and building their own platforms when existing ones shut them out.
Context knowledge is part of the technical work
My idea of Artful Intelligence holds that the best AI comes from combining technical skill with a close understanding of the people it serves. In the Caribbean that means knowing the informal economy, extended family structures, the religious and cultural frames people decide within, the region's disease burden, Caribbean English and Creole, and the racial and ethnic history of each society.
Many of the women I have worked with bring that knowledge to AI projects in depth. It is often described as a soft extra added to the real engineering. For AI that has to work in the Caribbean, it is part of the engineering: a credit model or health tool built without it rests on a wrong picture of how people live, and it fails in ways the test data never shows.
What the women on my teams taught me
Women who worked with me in AI training, assessment, operations and communications taught me things I would not otherwise have learned:
- how AI tools misjudge Caribbean women's financial profiles, for example when income comes from several informal sources;
- how AI training material assumes a North American workplace that excludes Caribbean professionals;
- what it costs, day after day, to keep showing up in technology spaces not built for you.
They also built programmes that reached people nobody had reached before, and AI ethics frameworks drawn from Caribbean values.
What men in AI leadership owe
Writing this as a man means looking at my own role in the structures I am criticising. Caribbean men who lead AI organisations, sit on advisory bodies, speak at events and shape AI education have concrete duties: recruit women into leadership roles, share platforms, give credit where it is due, and accept discomfort when current arrangements favour men.
I have not always met that standard. I am naming it as the one I hold myself to and the one I think Caribbean male AI leaders should hold each other to. International Women's Day is for recognising women's achievements, and for men it is a day to be honest about the work still to do.
What a fair Caribbean AI sector would look like
The institutions, funding, governance rules, talent pipelines and norms of Caribbean AI are being set now, which is when choices about gender have the most effect. The sector I want to see has:
- women leading AI organisations, not only working in them;
- women drafting national AI strategies, not consulted at the end;
- women's names on AI research from UWI, UTech, the University of Guyana and other Caribbean institutions;
- Caribbean women recognised internationally as AI experts;
- girls who see women in AI leadership and assume the field is open to them.
Without deliberate choices, Caribbean AI will copy the global pattern of male dominance and build systems that work worse for Caribbean women. I do not know yet which institutions will act first, and I suspect funding bodies matter more than training programmes, because money decides who gets to lead.
To the Caribbean women building AI without waiting for permission: StarApple AI sees your work, will back it, and wants you in this field.
What to do next
- Women starting in AI: pick one problem from your own field, such as a clinic's appointment backlog or a school's reporting load, and build a small AI-assisted fix within three months. A working example counts for more than a certificate.
- Heads of AI organisations: publish the gender split of your leadership, speakers and project leads once a year, and set a target for the next.
- Event organisers: do not confirm a panel on AI until at least half the speakers you have invited are women with technical or policy roles.
- Funders and investors: track what share of your AI funding goes to women-led teams and report it alongside your portfolio results.
- Men on AI advisory bodies: when you are invited to speak, recommend a woman colleague for at least one invitation in three.
Frequently Asked Questions
Do women need to learn to code to work in AI?
No. Many AI roles depend more on judgement and domain knowledge than on programming: testing and evaluating AI outputs, AI product management, data governance, policy, and training others. Coding helps, and basic Python is worth learning eventually, but a nurse, lawyer or accountant who understands where AI fails in her field is already valuable. Start from your field and add technical skills as the work needs them.
Where can Caribbean women find AI peers and mentors?
Global networks such as Women in AI and Women in Machine Learning run online events and mentoring open to members anywhere, including the Caribbean. Regionally, university computing departments, national AI and ICT associations and StarApple AI's free sessions are places to meet practitioners. Asking one experienced person for a monthly half-hour call is often more useful than joining several groups.
How can an employer check whether its AI hiring tools are biased against women?
Compare selection rates for women and men at each stage the tool touches, such as CV screening and interview shortlists. A common rule of thumb from US employment practice flags a problem when one group's selection rate is below four-fifths of the other's. Ask the vendor what data the tool was trained on and whether it has been tested on candidates from your country, and keep a person reviewing every rejection until you have checked.
How can women take part in national AI policy?
Watch for public consultations from the ministry responsible for technology and from national AI task forces, and send written comments before the deadline. Submissions from professional bodies usually carry more weight than individual ones, so working through your association helps. Offering specific evidence from your own sector, such as where an AI system failed a client, is the most persuasive contribution.
How can men in AI leadership tell whether they are sharing platforms fairly?
Count. Keep a simple record for a year of the speaking invitations, media requests and board seats offered to you, and how many you passed to women colleagues. Do the same for who leads your projects and whose names appear on published work. The numbers usually differ from what people assume, and they give a baseline to improve on.
If you are a Caribbean woman who wants to build AI skills, StarApple AI's training programmes are open to you. Connect with us at insights@starapple.ai.
Join StarApple AI