PRECISION-AI: Chile's 5 billion-peso bet on AI oncology, and why the real challenge is clinical adoption

By University of Chile · July 8, 2026.

The University of Chile and Universidad del Desarrollo have unveiled PRECISION-AI, a technology program awarded by Corfo with 5 billion Chilean pesos for five years of execution, aimed at developing artificial intelligence solutions that help prevent, diagnose, treat and prognose cancer. The project is co-directed by Dr. Arnaldo Marín, of the Basic Clinical Oncology Department of the University of Chile's Faculty of Medicine, together with Dr. Ricardo Armisén on behalf of UDD, and involves the University of Chile Clinical Hospital, Hospital del Salvador and a network that brings together public, private and international institutions, including the backing of MIT and the London AI Centre for Value-Based Healthcare. One of its most relevant components is the construction of a federated, secure and ethically designed data infrastructure, intended to train models without exposing patients' privacy, and with the aim of later becoming available to the entire national and international ecosystem.

The context justifying the investment is harsh: cancer is the second leading cause of death in the world and the first in Chile, with close to 60,000 new cases a year and some 30,000 deaths, according to figures cited by UDD rector Federico Valdés at the presentation. Health Minister May Chomalí offered a figure that, broadly speaking, says more about the sector than the funding amount does: of the six technology programs Corfo awarded in the AI-for-Chilean-industry call, three went to health. That confirms that precision oncology —cross-referencing genomics, imaging and clinical data with AI models to tailor diagnosis and treatment to each patient— has become the terrain where governments and academia worldwide are concentrating their bets, because it is where the return in lives is most tangible and where basic science already has decades of groundwork waiting for a compute layer to make it scalable.

That said, it is worth being honest about what this announcement is and what it is not yet. PRECISION-AI is, at this stage, a governance and funding framework to «organize and articulate» —in Dr. Marín's own words— existing research projects, not a clinical product ready to be used in a hospital. And Minister Chomalí herself put her finger on the most honest sore point of the whole day: the barrier is not only technological, but one of medical and legal responsibility. «Who is going to be responsible if it's the AI that's looking at the image? How am I going to take charge of the results?» she asked, and added the most common risk of this kind of initiative in public health: getting stuck as an eternal pilot program without becoming persistently integrated into the care process. It is exactly the pattern we have seen repeated in dozens of health AI projects around the world —good science, good initial funding, but regulatory and organizational friction that prevents the leap into daily practice—. That the health authority itself says so at the presentation, rather than selling the project without nuance, is an unusual sign of maturity and worth acknowledging.

Our reading is that PRECISION-AI's value lies not so much in the amount —modest compared with what the major cancer centers or AI labs move globally— as in the architecture it proposes: a federated, shared data infrastructure is, potentially, more valuable in the long run than any single model, because it is the asset that allows the next generation of diagnostic tools to be trained and validated without each hospital having to solve the data-privacy problem from scratch. It is the same pattern we are seeing in other geographies: the dispute is no longer only about who has the best model, but about who controls the data and clinical-integration layer on which those models operate. If Chile manages to make that infrastructure real, interoperable and effectively used —and not a promise that dissolves once the Corfo funding ends—, the country positions itself with a digital health-sovereignty asset that is uncommon in the region.

This connects to the underlying thesis we hold at Zendoric: AI-assisted precision oncology is exactly the kind of application where long-term optimism has solid foundations. Cross-referencing genomics, clinical history and imaging with models capable of detecting patterns that would take a human eye years to identify is, literally, part of the path toward reducing cancer from a death sentence to a manageable chronic disease, and over time toward its practical eradication as a leading cause of death. But that horizon does not arrive by decree or by budget: it arrives when projects like this one solve, one by one, the boring and difficult problems of clinical governance, legal responsibility and hospital adoption that Minister Chomalí named without mincing words. Chilean talent and basic science, endorsed here by MIT's backing, are not the bottleneck; translation to the patient's bedside is, and that is where this program will have to prove, over the next five years, whether it lived up to its own diagnosis.