Latin America Plans Regional AI Model Launch September

A coalition of over a dozen Latin American countries is set to launch Latam-GPT in September 2025, the region’s first large-scale AI language model tailored to local cultures, languages, and Indigenous dialects such as Rapa Nui. Led by Chile's National Center for Artificial Intelligence (CENIA) and supported by more than 30 regional institutions, the open-source project aims to democratize AI usage across Latin America by addressing the biases and limitations of predominantly English-trained global AI models. Based on Meta’s Llama 3 architecture and trained via a distributed network of regional computers, Latam-GPT is intended as foundational technology for applications like chatbots, virtual public service assistants, and personalized educational tools, rather than a direct competitor to consumer-facing AI products like ChatGPT. Chilean Science Minister Aisen Etcheverry highlighted the model’s potential to revolutionize AI adoption in schools and hospitals by reflecting local culture and language. Although the project currently lacks a dedicated budget, support from the regional development bank CAF and Amazon Web Services, along with hopes to attract further funding through demonstrated capabilities, underscores a growing push for culturally inclusive AI development in Latin America. This initiative marks a significant step toward creating AI technologies that better serve the Global South by preserving Indigenous languages and integrating regional cultural nuances.