This is an excerpt from The Future of Fintech in Latin America 2023 report.
Latin America has seen a shift in its financial payments industry in recent years, with countries such as Brazil and Mexico morphing into fintech powerhouses. AI initiatives are being adopted in public and private sectors to facilitate use of new technologies
in Latin America. The region is also seeing competition in the market, as some nations have adopted AI faster than others.
According to Latin America Reports, seven in 10 people in Latin America are unbanked or underbanked, but the number of fintech companies in the region has doubled since 2020. The Covid-19 pandemic catalysed the digital payments shift in the region, with
13 million people making their first online transactions in 2020.
AI is expected to increase the GDP of Latin America by more than 5% by 2030, according to an Economist Impact report, as demonstrated in the graph below. However, according to Oxford Insights, Latin America is among the lowest scoring regions for the Government
AI Readiness Index in 2021.
Latin American fintechs are adopting AI technology for speed and security
The financial sector has seen the most investment compared to all other industries in Latin America, with fintech firms receiving 40% of all investment capital. This has been reflected in the growth of fintech unicorns in Brazil, Argentina, and Mexico, such
as Nubank and Konfio.
Gabriel Siler, data director for BBVA Mexico, remarks that Latin America is priming itself for AI adoption by investing in AI research, developing AIbased solutions, and building technological infrastructure. Siler cites BBVA’s collaboration with Universidad
Panamericana to create a data science Master’s Degree programme for BBVA employees as a move in the right direction.
AI is largely being implemented in healthcare in Latin America, with an estimated growth in healthtech since 2019. Another major industry in the region in agriculture, which could see AI improve application of sustainable initiatives by agritech startups,
however the sector has seen little investment in recent years. Siler adds that Latin American fintechs are focusing on payments and remittances, fraud prevention, credit scoring, lending, and insurance.
Chief data officer at Mexico-based BNPL and consumer-lending company Kueski, Krishna Venkatraman, notes that open banking is a big focus for AI initiatives to help the underbanked access financial services, and states that Kueski’s main aim in AI usage is
to boost accessibility to the unbanked and underbanked consumers in Mexico: “AI allows us to make faster, better, less expensive decisions through the use of intelligent data, ultimately driving costs down and further opening up the market for consumers.”
Brazil is one of the largest areas of growth for fintech development, which predisposes the country to have a high rate of AI adoption. 7 out of 10 companies in Brazil are deploying AI strategies in their businesses, mainly concentrated on enhancing user
experience.
Alexandre Magnani, CEO of PagBank PagSeguro, observes that AI is enhancing accuracy and security in the Brazilian payments ecosystem: “AI-powered technologies such as machine learning and natural language processing are helping to automate processes and
reduce costs. Additionally, AI is enabling more accurate real-time fraud detection, as well as personalised services and customer experiences. AI is also being used to optimise payments routing, improve customer segmentation and targeting, and enable more
accurate and efficient customer service.”
Fintechs such as Mexico City’s Finvero, a consumer-leading marketplace; N5, a platform that aids legacy platforms in their digital transition which was founded in Argentina; and Brazil’s fraud protection company, ClearSale, are a few of Latin America’s leading
fintechs, driving the region towards a digital economy. All three companies use AI from Microsoft Azure to drive their businesses forward.
National strategies indicate a push for AI
National AI strategies in the Latin American region are mainly focused on developing talent, building on technological infrastructure, and ensuring the use of responsible and ethical AI. However, national strategies are dependent on political administrations
in power and their priorities.
Multiple nations in the region have implemented national AI strategies, including Argentina, Brazil, Chile, Colombia, and Uruguay. Both Brazil and Colombia are globally ranked in assessments of AI strategies, with Colombia scoring higher than the US and
Germany. Colombia’s framework prioritises inclusivity in AI algorithms and the introduction of regulatory sandboxes.
Magnani states that Brazil is prioritising an AI-driven economy and developing technology infrastructure: “The government is investing in the development of AI-related research, development, and innovation initiatives, as well as providing incentives for
businesses and universities to adopt AI technologies. Furthermore, Brazil is prioritising the development of skills and expertise in AI through training programs and the promotion of AI research and development.”
Brazil’s strategy aims to include digital literacy and AI courses in the national curriculum, and Chile’s approach concentrates of data governance and modernising laws on technology. However the new political organisation in control in Chile might discontinue
aspects of their AI strategy.
In Argentina, the administration under Alberto Fernandez halted developments of a national AI framework. Mexico has not established a national AI framework, despite interest in pursuing an official AI approach in the previous administration.
Ethical AI and data privacy are top priorities for Latin American regulation
Regulation is moving at different speeds in various countries in the region, with some more focused on building on the foundation for AI adoption than others. Currently, Latin American countries are focusing on data privacy and ethical AI legislation independently.
Data privacy regulation is an issue that many Latin American nations are focused on solving when it comes to AI, however there is no unity in data protection regulation in the region. Countries such as Mexico, Columbia, Peru, and the Dominican Republic have
had data protection legislation in place since the early 2010s, whereas other nations, such as Brazil and Panama, have only enacted similar regulations in the last five years.
Working to solve inconsistent regional data protection regulation is the Economic Commission for Latin America and the Caribbean (ECLAC), which has laid out principles for data protection standards and transparency guidelines.
The Association for Civil Rights in Argentina has established a framework for inclusion of human rights and strengthening privacy protections in AI, and have outlined the following: "Adopt international human rights standards as the main framework through
which to assess the effects of AI on human beings. In this context, ethics can play a role but must be complementary to the human rights approach. Thus, the paradigm of ‘ethics by design’ and ‘privacy by design and default’ should be incorporated into any
AI initiative that uses personal data."
“Ethical AI is being implemented in the payments industry in Brazil in order to ensure that AI-powered technologies are being used responsibly and in accordance with ethical principles. This includes the development of regulations and standards to protect
data privacy, as well as the development of responsible AI principles to guide the development and use of AI technologies. Additionally, the industry is taking steps to ensure that AIpowered technologies are transparent and accountable and are being used in
a way that benefits society as a whole,” states Magnani.
Overall, AI regulation is lacking standardisation in Latin American countries. Despite a widespread interest in driving national AI strategies, regulatory authorities are still in process of creating guidelines and principles for how AI should be integrated
in the region’s fintech ecosystem.
AI bias and lack of resources threaten AI development in the region
Challenges to the widespread adoption of AI in the region include low access of education, limited resources to develop complete and unbiases AI algorithms, and the impact of outside technology in Latin American nations. Governments in South America have
experimented with AI devices, such as an AI tool designed to classify rulings in the Columbian Constitutional Court and an algorithm that creates a risk index for youths in Chile, both facing issues of accuracy and bias within the technology.
The paper ‘Ethical Considerations of AI in Latin America’, describes the use of Western technology as a tool of colonialism from which Latin America will suffer due to the promotion of specific values and applications that facilitate data extraction for
the benefit of these outside providers. They state the following: “With regard to technology, research and production are disproportionally concentrated in the Global North, but its products and applications are deployed globally. Thus, technology functions
as a tool for coloniality, exacerbating the power differential between technology developers and users. Other scholars have referred to this phenomenon as digital colonialism or algorithmic coloniality.”
While this is a major concern in terms of outsider third-party providers providing technology that cannot achieve national priorities, this can be avoided through widespread promotion of digital and AI literacy, more Latin American sourced talent, and ethical
AI models that work towards solving bias in AI.
“When it comes to ethical AI, we are constantly working to ensure that the different attributes we use to make customer decisions are not based on any bias (i.e. gender, age, race, etc.) AI can easily be discriminatory if not used ethically and with a conscious
eye on avoiding bias. We use AI to quickly analyse customer data based on credit or fraud risk factors – factors that are independent of protected categories,” says Venkatraman.
Siler touches on the need for more resources to keep up with global AI providers: “For SMEs and even larger companies, without the right infrastructure in place, they may struggle to store and manage the massive amounts of data required for AI applications
and face limitations in terms of processing power and scalability. Elastic cloud services could help to get a scalable and flexible infrastructure for storing and processing data, which is essential for successful AI applications. It's important to keep in
mind that elastic cloud services to support AI applications require a significant upfront cost, and companies of all sizes should justify this cost in terms of the opportunity cost of not investing, which may eventually place them out of the game.”
Latin America is in process of building its technological infrastructure for widespread AI adoption. Fintech leaders in the region are committed to confronting issues of talent, outsider bias, and the significant unbanked and underbanked populations. Backed
by governments developing national AI strategies, the region is primed to become a powerhouse in the fintech sector in coming years.