Stanford HAI Names 15 PhD Researchers to New Data Science Scholars Cohort

There's a quiet but important question running underneath a lot of today's AI anxiety: who is being trained to actually understand this technology well enough to guide it responsibly? Not just build it faster, but study its effects, catch its failures, and translate its possibilities into things like better cancer imaging or more resilient power grids. Stanford HAI's latest answer arrived this week with the announcement of 15 new Data Science Scholars, PhD students selected for a two-year fellowship meant to grow exactly that kind of talent.

The Stanford HAI Data Science Scholars Program draws from existing PhD students across all seven of Stanford's schools. It's not limited to computer science departments, which is part of the point. Scholars are doing foundational work on methods, models, and theory, but also applied research in fields like scientific discovery, health and biomedicine, mental health, language and culture, education, and the arts. A third strand looks at the ethical, societal, and policy consequences of these tools, the part of the AI conversation that often gets the least funding and the most urgent need.

Selection criteria favor the significance of a scholar's research questions, the rigor of their methods, and a demonstrated willingness to work across disciplines. That last point matters more than it might sound. AI research has a habit of staying siloed, with computer scientists building tools that never quite reach the clinicians, lawyers, or sociologists who most need them. This program is explicitly designed to counter that, building what HAI calls a "critical mass" of early-career researchers whose combined work spans scholarly discovery, educational change, and broader societal impact.

A Cohort That Spans Disciplines

This year's group reads almost like a cross-section of modern intellectual life. It includes scholars from computer science, statistics, law, neuroscience, sociology, education, physics, materials science, and business. Their projects range from AI governance and cyber conflict to protein design, radiology imaging models, and research into the structure of the universe itself.

"AI and data science are reshaping nearly every field of human inquiry, and the scholars we've selected this year reflect exactly the kind of breadth and depth that mission demands," said Stanford HAI Denning Director James Landay. "From AI governance to protein design, from radiology to the structure of the universe, this cohort is asking questions that matter, and bringing the rigor and creativity to answer them. We're proud to support their work at this critical moment."

Some examples help explain why that breadth is more than a talking point. Eve, a sociology PhD student, uses mobile GPS data and computer vision analysis of Google Street View images to study how economic inequality shows up in everyday city life, from who crosses paths with whom to which neighborhoods attract investment. Bruria, a JSD candidate at Stanford Law School and former intelligence officer, studies how cybersecurity firms shape international conflict and AI governance through both legal and computational analysis. Chelsea, an MD/PhD student in neuroscience, is mapping how ketamine disrupts the brain's internal models of space and emotion, work that connects directly to open questions in AI about how machines build and lose structured representations of the world.

Others are tackling problems with more immediate practical stakes. Jing, a statistics PhD candidate, is working with the National Laboratory of the Rockies on weather-aware electricity pricing designed to reduce grid stress during extreme weather, alongside personalized healthcare models that account for patient differences in risk prediction. Rohan, in biomedical informatics, builds machine learning methods to help medicinal chemists find drug candidates faster. Steven, a computational mathematics PhD candidate, leads an effort called Terminal-Bench-Science aimed at building standardized ways to evaluate AI research assistants on real scientific workflows, a kind of quality control system for AI doing science.

The cohort also includes researchers working at the intersection of AI and physical infrastructure. Luca, in materials science and engineering, develops models for materials discovery aimed at accelerating next-generation energy storage, while also building efficient machine learning tools for wearable health devices where battery life and privacy both matter. Victor, pursuing a joint JD/PhD in political science and law, focuses on AI governance and environmental sustainability together, a pairing that reflects how entangled these two policy domains have become.

What ties these projects together isn't a shared technical method. It's a shared recognition that data science has stopped being a niche skill and has become something closer to a universal research language, one that shows up in courtrooms, hospitals, classrooms, and physics labs alike.

Why It Matters

Programs like this one matter beyond the individual scholars involved. The AI workforce conversation tends to focus on engineers and companies, but the deeper, slower work of training researchers who can evaluate these systems, understand their limits, and apply them responsibly to real human problems gets far less attention and far less funding. A fellowship that funds a neuroscientist studying ketamine's effects on brain representations alongside a legal scholar studying cyberwarfare isn't just diversifying a research portfolio. It's building the kind of interdisciplinary talent pipeline that will be needed to keep AI's rapid expansion tethered to actual human benefit, in medicine, energy policy, education, and beyond. Whether that benefit reaches the people who need it most will depend less on any single algorithm and more on researchers like these, trained to ask not just what AI can do, but what it should do, and for whom.