by Jasmine Kendall

Thumbnail photo by Alex Brites from Pexels

In this latest instalment of – IAIDL’ AI strategy series, we turn our attention to the Latin American region to take a closer look at two neighbours’ approaches to artificial intelligence; Argentina and Uruguay. 

In this year’s Government AI Readiness Index, Uruguay topped the Latin American region, with an overall score of 55.57. Meanwhile, Argentina came in a regional fourth, behind Chile and Colombia, with an overall score of 50.75. As two regional leaders in this sphere, how these neighbouring governments tackle AI is likely to have an impact on the future digital and economic landscape in Latin America more widely.

There are a number of parallels between Argentina and Uruguay’s AI strategies. They were published only months apart in 2019, and both were designed by previous administrations which have since been defeated in national elections. Yet, as we explore in this blog, each is underpinned by seemingly very different priorities. For each area considered, we have determined which country’s strategy has the comparative advantage according to our analysis.

Argentina has a broader AI strategy, which was published in 2019 under the government of then-president Mauricio Macri. The document seeks to promote AI in the private sector, minimize ethical risks, and develop talent, amongst other objectives. Uruguay’s AI’s strategy, also published in 2019 under the administration of a former president, Tabaré Vázquez, is much briefer, dealing specifically with promoting the use of AI within public administration. Nonetheless, Uruguay’s Agenda Digital 2020, another output of the previous government, takes a much broader view of digital transformation in the country and is in many ways more comparable to Argentina’s strategy in this sense. As such, this blog also considers the documents, reports, and agendas which accompany the AI strategies. In doing this we hope to achieve a fairer comparison, which accounts for discrepancies between the ways in which each country structures its strategic documents.

Background

Before diving into the analysis of each strategy, it’s worth acknowledging their political backdrops;

As of May 2020, a report published by the Banco Interamericano de Desarrollo (BID) claimed that Argentina’s AI strategy can be classed as “to be continued”; i.e. not currently implemented. In October, Mauricio Macri’s incumbent conservative government was defeated by Alberto Fernández’s left-wing coalition Frente de Todos. Although the Fernández administration did initially show some willingness to commit to digital transformation by vowing to re-establish the Argentine Ministry of Science and Technologies – which had been demoted to a secretariat under the previous administration – Fernández’s mandate has since been dominated by the coronavirus response.

During its presidential elections in 2019, Uruguay experienced something of a reverse of the Argentine situation, when the incumbent socialist government was replaced by a centre-right President, Luis Lacalle Pou. Pou’s presidency seems to have run more smoothly than his Argentine counterpart’s to date; as opposed to Argentina, which was hit relatively hard by the virus, registering more than 44,000 deaths from COVID-19, Uruguay has only recorded 231 deaths at the time of writing. Digitalisation remains a government priority, and unlike Argentina, the BID considers Uruguay’s strategy to be a completed document which is currently in implementation. Uruguay immediately looked to technology in its virus response, creating online resources through which people could seek advice and report potential cases. This caught the attention of Apple and Google, and subsequently, Uruguay was the first country in Latin America to implement their track and trace technology.

Whilst it’s possible that the Argentine government may disregard or modify the 2019 strategy, it’s nonetheless the most up-to-date vision of artificial intelligence in Argentina, and therefore our main point of reference for this blog.

Vision: draw

Both the Argentinian and the Uruguayan AI strategies set out ambitious goals for a digital future, and received full marks in the vision dimension of our index. Nonetheless, the visions projected by each strategy have somewhat different focuses. Each strategy’s key objectives are summarised below;

 

 

Argentina’s AI strategy imagines a much more commercially focused approach, whereas its Uruguayan counterpart places much more emphasis on improving the use of AI in government. These divergent priorities are reflected in our Index scores. As explored further below, Argentina scored higher public sector readiness indicators.

Creating a market for AI growth: Argentina (for now)

According to our index, last year Argentina outperformed Uruguay when it comes to AI readiness in the tech sector. This is based on data which indicates that Argentina has over 6000 tech start-ups, compared to only 185 in Uruguay. Similarly, an analysis conducted by the Banco Interamericano de Desarollo also concludes that Argentina has a stronger entrepreneurial ecosystem, since it is home to several technology unicorns, such as Mercado Libre (a Latin American equivalent to eBay), Despegar (a travel booking website), or Globant (an IT and software development company).

Argentina’s fluctuating economy is no stranger to high levels of inflation and economic crashes, meaning the country is sometimes regarded as a difficult business environment in which to operate. Nonetheless, the 2019 AI strategy places significant emphasis on easing the path for tech giants and startups alike, potentially contributing to Argentina’s success in this dimension. The strategy references several initiatives in this regard, including;

• The creation of the National Fund of Entrepreneurial Capital (FONDCE) to support start-ups.

• Changes in legislation which make it possible to start a business in 24 hours.

• Tax benefits for individuals who invest in startups and SMEs.

Meanwhile, the Uruguayan strategy makes little mention of the economy, given its focus on public administration. The Agenda Digital 2020 is slightly more commercially orientated, but focuses on “sustainable economic development” and reducing the economic risks associated with AI, adopting a much safer tone than its growth-led Argentine equivalent. 

Despite this, many have noted that the Uruguayan tech sector is thriving, with new unicorns like dLocal poaching executives from companies based in Argentina, particularly since the Fernández administration introduced more interventionist government controls on the economy. Therefore, whether Argentina will be able to hold onto this lead in the coming year remains to be seen.

Public sector readiness: Uruguay

Whereas the Argentine strategy performs well under an economically focused lens, the Uruguayan government outstrips Argentina when it comes to the government indicators in our index.

Argentina’s AI strategy does propose some government-centric reforms, stressing the need for the digitalisation of government services, and highlighting existing initiatives, such as Boti, a virtual assistant which allows citizens to interact with local government in Buenos Aires. 

However, public sector reform hardly shines through as a central focus of the strategy, as in the case with Uruguayan AI strategy, which is primarily focused on AI reforms in the public sector. From the outset, the Uruguayan strategy explicitly sets out to “consolidate a closer relationship between the people and the state” and “provide more efficient and innovative services” to citizens, by increasing digital capacity within government. To achieve this, the Uruguayan strategy proposes to:

• Create a training programme around the use of AI in public services;

• Train all government departments using the aforementioned programme;

• Define standards, guidelines, and recommendations for auditing of the decision-making algorithms used in government; amongst other initiatives.

Feasibility and implementation: Uruguay

Having a high-reaching vision for change is one thing, but as our initial ‘What makes a good AI strategy?’ blog sets out, AI strategies need to set out measurable goals if they are to translate ambitious ideas into practical change.

This is something the Uruguayan strategy and its accompanying initiatives do particularly well. Uruguay Digital’s website allows visitors to search for particular digital initiatives proposed in the Agenda Digital 2020, and see how close they are to full implementation, measured in percentage points. Elsewhere, AGESIC, the ministry for digital government and society responsible for creating Uruguay’s AI strategy, places significant emphasis on ‘actually making things happen’ and ‘improving the improvable’ – suggesting that the Uruguayan government’s digital departments have considered the importance of a culture of tangible change.

Example progress markers on Uruguay Digital’s website. Source: https://uruguaydigital.gub.uy. 

The Argentine government lacks a similar tool, making it hard to track progress on some of the claims made in the AI strategy. To give credit where it’s due, many of the goals proposed in the strategy do have clear markers. For instance, the previous Argentine government sought to measure their investment in human capital and digital education by monitoring tracking the number of degrees related to AI available, scholarship programmes abroad, and papers published, amongst other indicators. Yet the government does not seem to have published any data surrounding whether these goals are being met, likely due to institutional discontinuity.

Final takeaways

Many question marks hang over both Argentina and Uruguay’s AI strategies. Are they simply vestiges of old administrations, as the apparent abandonment of Argentina’s strategy seems to suggest? How will each country’s new government address the challenges and opportunities associated with AI? 

To date, it seems like Luis Lacalle Pou’s relatively new Uruguayan administration has been quicker to take up the baton on AI than its Argentine equivalent, integrating AI into a coronavirus response which has been praised internationally. Meanwhile, following a change in administration, very little information is available at all surrounding the Fernández administration’s approach to AI; Argentina’s AI strategy, released by the ArgenIA group in 2019, has not yet been approved by the new government in a resolution.

As the Uruguayan case demonstrates, AI can be a powerful tool for governments coordinating a public health response to the virus. Yet, looking further forward, the digital sector will continue to be crucial for administrations as they seek to revitalise economies in the wake of the crisis. As inflation in Argentina nears 40% and the Argentina Central Bank runs low on dollars, how the Argentine government addresses AI in the coming years has the potential to define the nation’s economic trajectory. At – IAIDL, we’re eager to see how both administrations will make their mark on national AI strategies in Argentina and Uruguay as they approach one year in office.