How Generative AI Is Reshaping Work in Latin America Amidst Digital Inequality

The study examines how Generative AI (GenAI) impacts labor markets in Latin America, highlighting that wealthier, urban workers benefit more from AI's potential, while the digital divide limits gains for poorer, rural populations. Addressing this gap is critical to ensuring equitable AI-driven productivity across the region.

Researchers Paweł Gmyrek, Hernan Winkler, and Santiago Garganta from the International Labour Organization (ILO) and the World Bank examine the potential impact of Generative AI (GenAI) on labor markets in Latin America and the Caribbean (LAC). As most of the research on AI's effects has focused on high-income countries, this paper addresses the lack of knowledge surrounding how GenAI might shape the economic futures of developing regions like LAC. The study estimates the extent to which the labor markets in these countries are exposed to GenAI, using data from harmonized household and labor force surveys. The research takes into account the slow pace of technology adoption in the region and factors in the digital divide, which creates barriers to realizing productivity gains from AI tools. By focusing on the intersection of employment and access to digital infrastructure, the study sheds light on how the region’s economic inequalities may be exacerbated by the digital divide.

Unequal Exposure to GenAI: The Role of Location and Income

The findings show that certain sectors and demographics are more likely to experience high exposure to GenAI. Specifically, workers in urban settings with higher education, jobs in the formal sector, and higher incomes are most likely to engage with this technology. Younger workers are particularly exposed to the risk of job automation, with sectors such as finance, insurance, and public administration showing the greatest vulnerability. However, it is not only the risk of automation that is significant but also the opportunity for productivity enhancement through AI augmentation. Despite the potential benefits, the digital divide severely limits the extent to which these technologies can be applied. The study finds that nearly half of the jobs that could benefit from GenAI are hindered by inadequate access to digital tools. In poorer countries within the region, this digital divide is more pronounced, further limiting the potential for positive outcomes.

The Digital Divide: A Critical Barrier to AI’s Promise

One of the key aspects of the research is its analysis of the digital infrastructure and its role in determining whether GenAI will act as a buffer or a bottleneck for job creation and productivity in LAC. The report shows that between 30% and 40% of employment in the region is exposed to GenAI, with significant variation between and within countries. The total exposure includes jobs that could either be automated, augmented by AI, or fall into an uncertain category where the future impact of AI is difficult to predict. A relatively small portion of the labor market, about 2% to 5%, is at immediate risk of full automation. While these numbers might seem low, they still represent a significant number of jobs, especially in vulnerable sectors. On the other hand, jobs with potential for augmentation, where AI could assist and enhance human workers rather than replace them, make up a larger share around 8% to 12% of total employment in the region. This augmentation potential is particularly high in sectors like education, health, and personal services, which are less vulnerable to automation but stand to benefit from AI-enabled productivity gains.

GenAI’s Benefits Hampered by Limited Digital Infrastructure

Despite the potential for GenAI to augment jobs and boost productivity, the lack of access to computers and digital infrastructure is a critical barrier. The research shows that in lower-income countries, a significant proportion of jobs that could be enhanced by GenAI are held back by insufficient access to technology. For instance, in countries like Nicaragua and Honduras, the digital gap is so wide that nearly half of the jobs exposed to automation or augmentation do not involve the use of computers. This lack of access means that many workers who could theoretically benefit from GenAI will miss out on the productivity gains that come with it. Women and younger workers are particularly affected by this gap, with 6.24% of jobs held by women and 6.22% of jobs held by men being hindered by the lack of digital infrastructure.

Wealthier Nations Experience Higher GenAI Exposure

The study also highlights how economic status plays a role in exposure to GenAI. Wealthier countries in the region, such as Uruguay and Costa Rica, have higher levels of overall exposure to AI technologies, with total exposure reaching up to 40% of employment. In contrast, poorer countries experience lower exposure rates, partially due to the slower adoption of digital tools. However, even in wealthier countries, the uneven distribution of digital infrastructure within regions means that rural areas and low-income workers are less likely to benefit from AI technologies. This creates a risk that GenAI could widen existing inequalities within and between countries in the LAC region.

Closing the Digital Divide: A Path Toward AI Equity

Overall, the paper provides a nuanced picture of how GenAI might shape the labor markets of Latin America and the Caribbean. While there are opportunities for productivity gains and job enhancement, particularly in sectors like education and health, these opportunities are unequally distributed. The digital divide remains a significant barrier to the widespread adoption of AI technologies, especially in poorer countries and rural areas. This divide is likely to limit the potential benefits of GenAI for many workers, exacerbating existing economic inequalities. The paper calls for policymakers to address these digital gaps to ensure that the potential benefits of GenAI can be realized across the region and not just in wealthier, urban areas. Without such interventions, the risk is that GenAI will act as a bottleneck rather than a buffer, reinforcing existing inequalities rather than alleviating them.