KINGSTON, Jamaica — Jamaica risks overlooking women’s talent and adopting artificial intelligence (AI) systems that fail to account for their needs unless more women help shape the technology, Spectrum Management Authority (SMA) Managing Director Dr Maria Myers-Hamilton has warned.

She is calling for greater participation by Jamaican and Caribbean women in developing, testing and governing AI, as its applications expand across recruitment, healthcare, financial services and other areas affecting everyday life.

“Jamaica and the wider Caribbean cannot afford talent leakage of any kind,” Myers-Hamilton said.

She argued that countries seeking to benefit from AI must address women’s recruitment, retention and advancement, ensuring that their involvement extends to leadership and decisions about how the technology is used.

The World Economic Forum and LinkedIn’s 2025 “Gender Parity in the Intelligent Age” report found that women accounted for 29.4 per cent of LinkedIn members listing AI engineering skills, up from 23.5 per cent in 2018. The figures reflect skills listed on the platform rather than the entire AI workforce.

Myers-Hamilton said the gap has implications beyond access to technology jobs. The experiences represented in development teams and datasets can influence which problems are identified, how systems are tested and whose needs go unaddressed.

“Representation in AI is therefore not a diversity metric grafted onto engineering; it is an engineering control,” she said.

She pointed to the 2018 “Gender Shades” study, in which researchers Joy Buolamwini and Timnit Gebru tested three commercial systems that classified gender from facial images. Error rates reached 34.7 per cent for darker-skinned women, while the maximum error rate for lighter-skinned men was 0.8 per cent.

The findings applied to the systems examined at the time, but demonstrated how overall accuracy could conceal substantial differences in performance across demographic groups.

Myers-Hamilton noted that women researchers helped expose those disparities, underscoring the importance of different perspectives in identifying weaknesses and scrutinising technology.

For Caribbean women, she identified a further challenge: systems developed overseas may not adequately reflect the circumstances of people in small, developing countries.

“Women in the Global South stand at the intersection of both exclusions: under-represented in AI as women, and under-represented in AI’s training data and design assumptions as citizens of small and developing states,” she said.

The issue also carries economic significance for Jamaica, where women have formed a substantial share of the global services workforce. Estimates from the Global Services Association of Jamaica, reported in March 2023, put sector employment at approximately 59,000 as of February that year, with women accounting for around 71 per cent.

A May 2025 assessment by the International Labour Organization and NASK identified clerical occupations as having the highest exposure to generative AI. The researchers stressed that exposure does not necessarily mean job losses, with the transformation of jobs, rather than their replacement, considered the most likely outcome.

Myers-Hamilton said access to skills and opportunities must be supported by reliable communications infrastructure and affordable devices.

“Connectivity is the precondition,” she said. “Meaningful access to AI presupposes meaningful access to spectrum, networks and affordable devices.”

She identified regional cooperation, governance frameworks and networks supporting women in technology and digital policy as mechanisms for strengthening Caribbean participation.

Women’s involvement, she maintained, must span the AI life cycle: identifying the problems systems should address, selecting data, testing performance and establishing safeguards.