Sheinbaum outlines Mexico AI and supercomputing push in report

Mexico is sketching out its own version of an AI industrial strategy, and Sheinbaum’s report makes clear it is not just about chips or one flagship model. The government is tying together a national supercomputing center, semiconductor design initiatives, open‑source software factories and free AI training programs for young developers. Partnering with the Barcelona Supercomputing Center on the Coatlicue supercomputer gives Mexico access to European expertise while keeping computation and talent development at home.([es.wired.com](https://es.wired.com/articulos/chips-ia-y-autos-electricos-los-proyectos-tecnologicos-de-sheinbaum-para-mexico))

For the AGI race, this underscores that middle‑income countries are no longer just markets or data sources; they are building their own infrastructure and talent pipelines. A domestic AI workforce trained in public institutions and deployed through a “Fábrica de Inteligencia Artificial” for government use could give Mexico more leverage in negotiating with US and Chinese vendors, and reduce its dependence on foreign clouds for sensitive workloads.

If successful, this model could be copied across Latin America: combine public cloud‑adjacent supercomputing, open‑source tooling and large, subsidized training cohorts. That will not produce frontier AGI labs overnight, but it will broaden the base of people and institutions capable of contributing to advanced AI projects, which in aggregate nudges the global capacity for AGI‑relevant work upward.