Olivier Elemento, Director of Englander Institute for Precision Medicine at Weill Cornell Medicine, shared a post on LinkedIn:
“Medical Schools Need Departments of AI
Every major medical school has departments of Biochemistry, Pharmacology, Pathology. These disciplines became foundational to medicine, and institutions built homes for them. Molecular biology was once considered ‘just a method’—until it became so central that it needed its own departments.
I think AI is at that inflection point now. Yet most medical schools have no Department of AI.
I’ve written before about the three pillars of any AI initiative: data, people, and compute. A Department of AI gives institutions the structure to deliver all three—plus something more.
Research across the spectrum
AI is transforming both clinical medicine and basic research. In the clinic, AI powers diagnostic imaging, decision support, and risk stratification. In the lab, it enables protein structure prediction, drug discovery, and single-cell analysis. A Department of AI serves both—providing methods expertise from bench to bedside. And it can contribute directly to clinical operations and revenue, not just research. Academic medical centers have clinical data and domain expertise that industry lacks.
Data access and governance
AI runs on data. As I wrote recently, the biggest bottleneck is often just getting the data. Individual labs shouldn’t need to build datasets—that should be institutional. Faculty scattered across departments can’t advocate collectively. A department creates unified voice to push for data access—internal data lakes, sharing agreements, FAIR-compliant repositories.
Compute infrastructure
Training foundation models requires GPUs—lots of them. Individual labs cannot justify this alone. A department provides shared compute and negotiates with cloud providers. Without institutional compute, researchers become users of industry models rather than developers of their own.
Training the next generation
A department ensures AI becomes part of the curriculum—for MD students, PhD students, and clinical fellows. It provides a home for faculty who specialize in biomedical AI and creates career paths that don’t exist when expertise is scattered.
Institutional gravity
Institutes and Centers are nimble—and rightly dissolvable when priorities shift. But if AI is foundational, it needs permanence. Institutes and Centers bridge strong departments—without an AI department, there’s nothing to bridge. Departments have tenure lines, budget authority, and hiring power that persist.
The value chain is clear: invest in AI infrastructure, enable research and clinical tools, improve patient care, generate revenue, reputation, and talent.
Yes, this requires real investment. Mount Sinai took the leap and created a Department of AI and Human Health. More should follow. I believe that schools that build and invest in Departments of AI now will shape how medicine uses AI for the next 20 years.”
More posts featuring Olivier Elemento.