Medellín, Colombia — Italy became the first government to ban OpenAI from operating in its country due to concerns about how the company and its infamous ChatGPT app were processing citizen data, and whether these actions fell foul of General Data Protection Regulations (GDPR) protecting digital privacy.
While ChatGPT is now back in service for Italians, the furor surrounding generative AI has only gotten louder. Following an open moratorium from thousands of tech experts calling for a more cautious approach to the free development of powerful AI models and mounting concerns about how closely generative AI infringes on intellectual property, the so-called “godfather” of AI, Geoffrey Hinton, quit Google due to a collection of concerns around the technology.
But where do Latin American governments and policymakers stand when it comes to the future of generative AI?
Which countries have data protection regulations?
Italy’s recent ban on the use of ChatGPT was in direct relation to how citizens’ data is being used and protected. This was possible under the European Union’s GDPR legislation, which came into force in 2018, with many countries globally following suit in the subsequent years. For Latin America, there’s no singular approach to data protection laws but countries including Mexico, Colombia, Chile, Argentina, Peru, Ecuador and Costa Rica have since enacted some form of data protection legislation.
Interestingly, many countries in Latin America actually have something known as Habeas Data historically embedded in their constitutions. Unique to the region, habeas data has a special significance borne from a historical struggle against repressive regimes. While countries such as Venezuela don’t have formal data privacy laws in place, the habeas data principle could act as a leverage to accelerate discussion on data privacy and processing.
Yet these policy are in relation to how a person’s digital data can be used and handled by organizations, but such regulations don’t necessarily have a connection to the wider questions around the future of AI.
Who has a national AI strategy?
Across Latin America there are in fact a proliferation of national AI strategies, expert councils and policy initiatives already in place. This is largely down to the economic importance of AI and its ability to positively impact many industries. AI is forecast to boost the region’s GDP by over 5% by 2030, with projections likely to rise if governments introduce policies to build talent and
expand their digital infrastructure.
The first countries to publish their strategies were Argentina, Colombia and Uruguay, followed by Brazil and Chile in 2021. These strategies all emphasize as their top priorities cultivating local talent, strengthening the technological infrastructure and ensuring that AI is deployed in a responsible manner.
However, these frameworks have all come before the current questions thrown up by generative AI. So, is it likely that Latin American governments will adjust their AI policies in the near future?
Are adjustments to AI policy and data protection likely?
While specific restrictions haven’t yet been placed on the latest generative AI solutions, it seems that Latin American universities and research centers are leading the discussion on where the future may be heading for the region.
For example, in Brazil the Getulio Vargas Foundation held an important public debate on the ethics and governance of AI in organizations in March 2023. This comes in the wake of Brazil’s most recent AI regulations, issued in September 2022, came under criticism for being too vague.
And Mexican university Tec de Monterrey addresses the impact of new technologies like ChatGPT in a virtual forum. The event aimed to discuss how the technology could affect the daily activities of companies and potential legal implications for business operations in future.
What will this mean for innovation in the region?
When it comes to new AI policies, general purpose AI is very often part of this conversation. This includes language processing and image processing models that help to produce tools like ChatGPT. However, such legislation will likely affect the startup ecosystem as a whole given that 45% of startups consider their AI system to be general purpose AI.
Latin America as a region is gaining increasing prominence as a new hotspot for startups and tech entrepreneurs, producing 34 unicorns in 2021 alone. In the wake of advanced AI being put under the spotlight, many local founders are concerned that stricter policies may hamper innovation that’s vital to the region’s growth.
Jose Pino, the Colombian founder of the cryptosecurity startup Andro, believes that “people should be free to choose what data to share and what data not to share. The topic is not new, serious and controversial data breaches and incidents like Facebook’s with Cambridge Analytics have happened in the past, resulting in greater transparency into how data is used by Big Tech.
“Regulation can become excessive or restrictive, as in the case of the ChatGPT ban in Italy or the recent failed attempt to monopolize digital identity in Colombia’s National Development Plan Bill. Regulatory decision makers should uphold the rights of consumers by seeking greater transparency but not by restricting access to innovative tools such as artificial intelligence language models,” he continued.
And with regards to striking a balance between regulations and protections, Pino thinks “the focus should be more on the source of data acquisition, its protection and management rather than its use. In other words, regulation should seek to impact how sensitive data is acquired, protected and traded instead of limiting the innovations that are leveraged on it.”
This sentiment is reflected by the Mayor of Medellín, Daniel Quintero. In a recent interview highlighted his belief that AI is essential to the continued development of both the city and the country, and additional investment in the technology should be a priority for Colombia’s president.
The final word
Although Latin American governments have yet to make any formal announcements in relation to the latest advances in AI technology it seems clear that any decisions will likely need to balance the need to protect citizens data and opportunities in relation to the economic opportunities that new innovation stands to have across industries in the region.