Will Moss '26 was tired of watching policymakers scramble in the dark. As artificial intelligence accelerated into health care—reshaping diagnostics, treatment plans, and patient care itself—regulators, health systems, and researchers found themselves navigating a fractured landscape of competing rules, guidance documents, and voluntary frameworks. There was no map. So Moss built one.
The Health and AI Policy Index, or HAPI, is a public database that tracks health care AI legislation, executive actions, regulatory guidance, and voluntary frameworks across state, federal, and international jurisdictions. It represents the kind of infrastructure that seems obvious only in hindsight—a centralized, searchable resource for understanding how different corners of government and health care are actually attempting to govern artificial intelligence. The project, developed during Moss's time as a Brooks senior, has already drawn attention in academic circles; his research was recently published in npj Digital Medicine.
"Right now, policymakers, researchers, and health systems are all trying to navigate a rapidly changing regulatory environment," Moss explained. "I wanted to create a centralized resource that helps people follow what's happening across states in real time."
Why does this matter? Because there is no single comprehensive framework governing health care AI. Instead, stakeholders must pick their way through what Moss describes as a "complex patchwork of policies emerging from a wide range of regulators and institutions." One state might establish requirements for algorithmic transparency. Another might focus on liability. A federal agency might issue guidance. A hospital system might adopt internal standards. Meanwhile, the technology itself evolves faster than policy can follow. The risk is that important decisions about patient safety, provider accountability, and equity get made ad hoc, jurisdiction by jurisdiction, often without stakeholders even knowing what their neighbors are doing.
HAPI solves this by aggregating the sprawl. The platform's January 2026 analytic snapshot cataloged 240 distinct policies across jurisdictions. Users can explore these policies by jurisdiction, stakeholder group, and impact level, and each entry includes summaries, implementation considerations, and trend analyses. It's designed to be accessible to the people who actually need it: policymakers trying to draft regulations, researchers studying health AI governance, and health systems trying to implement technology responsibly.
Moss developed this work at the Windreich Department of Artificial Intelligence and Human Health at Mount Sinai, where he gained hands-on exposure to how health AI is evaluated, implemented, and governed in real clinical settings. He also drew on previous internships in federal and state government affairs, which gave him the policy analysis skills and stakeholder engagement experience necessary to understand the regulatory terrain he was mapping.
As AI systems become more embedded in clinical decision-making, the stakes of governance grow clearer. Policymakers will need stronger frameworks around transparency, oversight, safety, and accountability. But those frameworks won't emerge from engineers or companies alone—they will depend on the policies and governance systems built around these technologies.
"The future of health care AI won't just be determined by engineers or companies," Moss said. "It will also depend on the policies and governance systems we build around these technologies." With HAPI, he's provided a tool to help build those systems with open eyes.