Argentina most exposed to AI impact in Latin America, study finds

Uncertainty over how generative artificial intelligence (GenAI) might impact the labor market is a concern shared by most countries. Argentina is no exception. In fact, there are grounds for this concern. According to a recent study by the Economic and Social Council of Buenos Aires City (in Spanish, CESBA), it is the Latin American country with the highest level of occupational exposure to GenAI. CESBA is a non-governmental public body that drafts legislation, technical reports, and recommendations for the city’s executive and legislative branches. Based on data from the International Labor Organization (ILO), they found that on a scale of 0 to 1, the country has an exposure level of 0.292. By way of comparison, Chile has 0.279, followed by Brazil (0.276) and Mexico (0.261). In the Americas, it is surpassed only by Canada (0.322) and the United States (0.332). CESBA President Manuel Socias told the Herald that this is due to the fact that Argentina is the country with the highest urbanization rate in Latin America — 97% of its population currently lives in cities. This means that the bulk of its employment is urban, with the majority (78%) of jobs located in the service sector, the economic branch most susceptible to the impact of GenAI. And there is no region in Argentina where the exposure is higher than in Buenos Aires City. Buenos Aires leading the way The main focus of the CESBA study, called the Occupational Exposure Index to GenAI, is the impact of GenAI on Buenos Aires City. According to the authors, this is a pioneering study on a global scale, as no other city has conducted such an analysis. The data reveal that Buenos Aires’ exposure level is 0.353, well above the national average. The closest comparison is Singapore — a city-state — which has an exposure level of 0.366. “If we were to compare it with New York or London, [the city] would likely rank lower,” Socias acknowledged. The high exposure level for Buenos Aires is tied to two major factors. The first is that employment is concentrated in the service sector — 88% of the total. This is distributed across retail, healthcare, education, finance, and public administration, as well as professional and technical fields. The second stems from the city’s high level of education. “This matters because generative AI is particularly good at tasks that for years required a college education, such as writing, synthesizing, analyzing, or programming,” the report explained. Transformation in the service sector Among service companies, one of the sectors where adoption is fastest is the software and IT sector. According to a national survey on AI adoption in small and medium-sized enterprises in Argentina cited in the CESBA report, 85% of companies in that sector use at least one AI technology. Knowledge-based services (KBS) are already Argentina’s third-largest export sector, generating US$10 billion in exports. Leandro Mora Alfonsín, executive director of Argencon — the business chamber that represents KBS companies — told the Herald that the “traditional role” of the programmer is evolving “from manual coding toward the orchestration and supervision of AI agents.” When asked whether there had been an impact on the hiring of young people for junior/trainee positions within the sector, he stated that “the leading KBS export companies in Argentina are expanding their workforce.” He cited the example of JP Morgan, which is in the process of adding 1,000 new employees, the hiring drives by consulting firms EY and PwC over the past year, and Accenture’s opening of new offices in Salta, Córdoba, and Mendoza. The CESBA report noted that evidence to date shows that, globally, companies adopting AI are seeing growth in total employment. However, at the same time, they systematically reduce the proportion of employees with little tenure. “In the U.S., the U.K., Germany, India, Japan, and Brazil, the proportion of junior workers declines after adoption — in every case and in a sustained manner,” the study explained. Mora Alfonsín acknowledged that in an “early phase of technology adoption,” operational, repetitive, and low-complexity tasks, which have historically been performed by younger workers, are being taken over by AI tools. He also stated that there is a shift in demand toward skills involving “supervision and judgment.” One of the companies experiencing these changes is Baufest, an international provider of software and technology services. Leonardo Tocci, the company’s Data & Applied AI Practice Head, said that they do not see artificial intelligence leading to a decrease in demand for young talent. “What is beginning to change is the type of work people do when they join an organization and, with that, the way they build their first few years of experience,” he told the Herald. He explained that this alters what companies expect from entry-level candidates, since it is no longer simply a matter of hiring people to perform repetitive tasks. “The challenge is to keep pace with that change and find new ways to develop talent.” Bridging the gap between companies and unions For Socias, there is a dichotomous debate that must be overcome. “We must move beyond the techno-optimist narrative that technological development will solve all of humanity’s problems on its own, as well as the apocalyptic narrative of the end of work,” he explained. Socias went on to emphasize the importance of establishing a “common ground for discussion.” This involves determining the actual exposure of each task and, based on that, building a roadmap so that AI becomes an “opportunity to increase productivity, free up workers’ time, and improve incomes.” The goal, he added, is to ensure that the changes AI brings to the world of work do not unfold solely due to “the inherent inertia of technology.” Socias explained that the study is the first step in a three-stage plan. The second stage will seek to assess actual adoption through in-depth interviews with business associations, unions, and sector specialists. The third stage will involve bringing together companies, unions, and members of the Buenos Aires City government to facilitate dialogue, build consensus, and develop recommendations to guide that adoption. Socias noted that they are already engaging in discussions with the General Confederation of Labor (CGT) and ILO representatives in Argentina. “Leaders and unions are fully aware that this is an issue coming their way and that they cannot turn a blind eye,” he stated, adding that many unions have already set up their own artificial intelligence labs. Regarding the business sector, and while noting that companies have historically been “less open to social dialogue” than unions, he called discussions between them so far “positive.” “It’s a new experience for them to be invited to discuss this,” Socias acknowledged.