Chile has officially introduced Latam-GPT, a new open-source artificial intelligence model designed specifically for Latin America. The project aims to reduce bias in global AI systems that are largely shaped by US-based technology companies.
Latam-GPT was developed by the Chilean National Centre for Artificial Intelligence (CNCAI), known as Cenia. The model has been trained on millions of data points collected across Latin America to better reflect the region’s cultures, languages, and traditions.
President Gabriel Boric described the launch as a major step forward for the continent. “Thanks to Latam-GPT, we’re positioning the region as an active and sovereign player in the economy of the future,” he said.
“We’re at the table, we’re not on the menu,” he added.
Also Read: Malala becomes second Pakistani to be honored with portrait at Oxford
Chile’s Science Minister Aldo Valle said the program was created to address prejudices and generalizations about Latin American people and countries. He warned that simply using foreign AI systems could lead to the loss of local traditions and perspectives.
Unlike closed models such as ChatGPT or Google’s Gemini, Latam-GPT is open-source. This means programmers can adjust and customize it according to their needs.
Universities, foundations, libraries, government institutions, and civil society groups from Argentina, Brazil, Chile, Colombia, Ecuador, Mexico, Peru, and Uruguay contributed data to train the model.
Cenia director Alvaro Soto pointed out that global AI systems contain only a small amount of Latin American data. This limited input can lead to stereotypes. For example, ChatGPT has described a typical Chilean man as someone wearing a poncho with the Andes Mountains in the background.
Latam-GPT has been trained on more than eight terabytes of data, equal to millions of books. The entire project was developed with a budget of $550,000, funded mainly by the Development Bank of Latin America and Cenia’s own resources.
The first version was built using Amazon Web Services cloud technology. In the future, it will be trained on a supercomputer at the University of Tarapaca in northern Chile.
Currently, the model focuses mainly on Spanish and Portuguese content. However, developers plan to include Indigenous Latin American languages in the next stages.
Latam-GPT will be available free of charge for companies and public institutions. According to Soto, it could help hospitals facing “logistical problems or issues with the use of medical resources”.
Although experts say its small budget gives it “no chance” of competing directly with global tech giants, local entrepreneurs see strong potential.
Also Read: China introduces world’s largest eVTOL aircraft, named Matrix
Roberto Musso, a Chilean digital entrepreneur, plans to use Latam-GPT to build customer service systems for airlines and retailers. He said businesses are “very interested in having their users express themselves and receive responses in the local language.”
One of Latam-GPT’s key strengths is its ability to understand regional “slang, idioms, and even speech rate”, helping reduce misunderstandings and cultural bias.
Around the world, other regions are also building their own AI systems. Singapore launched SEA-LION for Southeast Asian languages, while Kenya developed UlizaLlama to support Swahili-speaking expectant mothers with health services.
With Latam-GPT, Chile hopes Latin America will no longer be just a consumer of AI technology but a creator shaping its own digital future.
