Mexico Unveils KAL to Boost Data Sovereignty, Local AI

Mexico announced the development of KAL, a national large language model (LLM) designed to reduce technological dependence on foreign AI providers and retain locally generated data. The project was presented at the IA+Mexico forum, organized by Cipre Holding and Nvidia.

Minister of Economy Marcelo Ebrard says the country is advancing toward a national AI ecosystem: “It can be done. In fact, we have to do it, and we are already on that track”.

The announcement responds to Mexico’s strategic objective of strengthening data sovereignty and establishing governance mechanisms for AI. The Mexican government intends to ensure that the data generated by local users remain within national borders, which could reduce regulatory, privacy, and intellectual property risks. Ebrard says that a national model allows for broader access to AI tools in universities, which can accelerate talent development and adoption across industries.

These priorities align with a global trend in which governments build sovereign digital infrastructure to support economic competitiveness and protect sensitive datasets.

"Digital government represents a new form of public management that goes beyond the simple digitization of services. It is a complete ecosystem where open data, transparency, and citizen participation come together to create a new model of governance," says the Centro México Digital.

KAL is being developed by Saptiva, a Mexico-based technology company. The model will integrate about 500,000 datasets to enable context-aware processing of locally relevant information, says Angel Cisneros, President, Saptiva. He adds that the goal is to create a system that does not translate but instead “thinks in Mexican,” which he describes as a linguistic and semantic framework aligned with local usage.

Cisneros notes that interactions with foreign LLMs result in data being transferred abroad, with limited visibility into how that information is used. A national model could mitigate these risks and support compliance with emerging regulatory frameworks on data protection and algorithmic transparency.

KAL is positioned to support Mexico’s transition from a consumer of advanced technologies to a developer of AI infrastructure. Potential applications extend across manufacturing, logistics, financial services, government services, and other sectors where localized models can support automation, compliance, and operational efficiency.

The development of KAL is expected to increase demand for high-performance computing capacity and stimulate collaboration among government, universities, and technology providers. Governance considerations will require federal authorities to establish standards for training, usage, auditing, and bias mitigation. These elements are consistent with international discussions on safe and transparent deployment of AI.