Skip to main content
v2026.11,986 entries · CC-BY 4.0

Editorial · CASRAI · AI and ML research outputs

Stanford HAI Awards $3.5 Million in 2026 Hoffman-Yee Grants to Seven AI Teams

Stanford HAI awarded $3.5 million across seven teams in its 2026 Hoffman-Yee grants to advance human-centered AI in medicine, science, law, and education.

Published 16 Nov 2026· 7 minute read

Ask CASRAI · free to try

Ask about this story

2 free questions a day, no account.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

CASRAI is the reference for research administration — bookmark it for the next question.

On November 16, 2026, the Stanford Institute for Human-Centered Artificial Intelligence announced $3.5 million in funding for its 2026 Hoffman-Yee Research Grants, awarding $500,000 each to seven interdisciplinary faculty teams developing AI systems designed to augment human expertise across medicine, scientific discovery, astronomy, legal administration, workforce training, global infectious disease diagnostics, and classroom education.

Seven Interdisciplinary Grants Targeting Human-Centered AI

The 2026 grant cycle marks the latest biennial cohort funded through the Hoffman-Yee Research Grants program, which philanthropists Reid Hoffman and Michelle Yee established at Stanford University in 2018. For the 2026 cohort, Google.org provided supplementary program funding alongside the foundational philanthropic endowment. The awards provide an initial tranche of $500,000 per project for year one, with high-performing teams eligible to compete for substantial multi-year follow-on funding in subsequent evaluations. With this round, the Hoffman-Yee program has distributed a cumulative $31 million across 25 interdisciplinary faculty teams since its launch.

Each funded proposal was required to align with Stanford HAI’s three core institutional pillars: investigating the human and societal impacts of artificial intelligence, creating technologies that augment rather than automate human capabilities, and developing machine learning architectures inspired by human cognition. The selected projects pair computer science researchers with specialists in medicine, law, astrophysics, education measurement, and global health, focusing on high-consequence domains where unaudited generative AI tools pose unacceptable risks of error, bias, or skill degradation.

The Seven 2026 Hoffman-Yee Research Grant Projects

The seven funded initiatives span seven distinct scientific and institutional challenges, each developing open or verifiable systems to assist professional decision-makers:

Project Title Lead Investigator Disciplinary Domain Core Technical and Operational Approach
Making Precision Medicine a Reality with Human-Centered AI Monica Lam (School of Engineering) Clinical Oncology and Biomedical Informatics Building “Scientific Researcher,” an open-source platform combining theorem provers and transparent reasoning to reproduce clinical trials and examine breast cancer outcomes in underrepresented populations.
AI That Teaches Itself to Advance Scientific Discovery Carlos Guestrin (Computer Science, SAIL) Physical Sciences and Molecular Discovery Developing self-improving foundation models that autonomously generate testable hypotheses, decompose scientific bottlenecks into subproblems, and refine internal capabilities alongside human domain experts.
A Co-Scientist for Time-Domain Astronomy Risa Wechsler (KIPAC) Astrophysics and Observational Astronomy Deploying multiagent AI systems to triage nightly alert streams from the Vera C. Rubin Observatory, preserve institutional memory across rotating research teams, and coordinate rapid follow-up observations.
Advancing Research, Reasoning, and Reform in Law and Policy Daniel E. Ho (Law, Political Science, HAI) Legal Informatics and Public Administration Engineering accountable legal reasoning systems to support attorneys and public agencies navigating statutory complexity in housing policy, regulatory rulemaking, and administrative benefits adjudication.
Collaborative Intelligence for Workforce Upskilling Diyi Yang (Computer Science) Labor Economics and Workplace Training Constructing TaskNet, an interactive AI-driven occupational ontology, and piloting “Upskilling Together” to evaluate whether task augmentation yields superior economic and productivity outcomes relative to job automation.
Re-architecting Human-Centered AI for Universal Diagnostics Manu Prakash (Bioengineering) Global Health and Diagnostic Microscopy Creating automated self-driving diagnostic microscopes and assembling one of the largest open-access multi-disease optical microscopy datasets as a global public good for malaria and tuberculosis detection.
Measuring the AI-Enhanced Classroom Sanmi Koyejo (Computer Science) Educational Psychometrics and AI Benchmarking Developing psychometric assessment frameworks to evaluate whether student learning gains achieved with AI pedagogical assistants persist when students work independently without technological support.

Funding Architecture and Program Background

The Hoffman-Yee Research Grants serve as Stanford HAI’s flagship internal incubator for large-scale, high-risk research initiatives that standard federal grant mechanisms struggle to accommodate. Traditional federal funding pathways, such as standard R01 grants from the National Institutes of Health or core directorate awards from the National Science Foundation, frequently require preliminary empirical data that interdisciplinary teams cannot produce without seed support. By committing $500,000 in immediate, flexible first-year funding, the program allows computer scientists, social scientists, and clinical researchers to establish functional collaborative pipelines before applying for larger institutional awards.

The program’s staged funding structure is specifically calibrated to promote durability. Following the initial one-year grant period, teams undergo rigorous milestone reviews evaluating methodological rigor, stakeholder engagement, and prototype viability. Top-tier teams can secure up to $2 million in multi-year follow-on funding over two years. In late 2025, Stanford HAI selected five teams from its 2024 cohort to receive this second-tier support across advanced biology, creative generative tools, and cognitive health diagnostics. The addition of Google.org as a co-funding partner for the 2026 cohort reflects growing philanthropic and industry coordination around human-in-the-loop AI methodologies.

Leadership and Stakeholder Reactions

Stanford HAI leadership framed the 2026 awards as an intentional departure from industry trends that prioritize workforce replacement over human agency. “This year’s grant winners are reimagining what AI can do for humans and society,” said Stanford HAI Denning Director James Landay. “Across medicine, science, law, and education, they are building novel systems to work alongside human expertise, instead of replacing it.”

Principal investigators stressed that human-centered AI requires structural accountability rather than blind trust in generative model outputs. Monica Lam, lead investigator for the precision oncology initiative, noted that “rather than replacing researchers, Scientific Researcher will serve as a reliable partner that helps organize evidence, analyzes complex data, and supports scientific decision-making.” Computer science professor Carlos Guestrin highlighted that replacing opaque general-purpose models with domain-specific, self-improving scientific reasoning systems enables researchers to tackle foundational problems in chemistry and materials science that remain intractable for standalone tools.

In governance and legal contexts, where administrative law demands strict procedural transparency, legal scholar Daniel E. Ho emphasized democratic accountability: “Our goal is to develop, assess, and evaluate AI systems that are not only powerful but also legally rigorous, transparent, and aligned with democratic values.” In bioengineering, Manu Prakash emphasized that the diagnostics project centers clinicians in low- and middle-income countries as co-creators of open data rather than passive recipients of Western technologies, ensuring tools address neglected tropical and infectious diseases effectively.

Implications for University Research Offices and AI Scholars

The 2026 Hoffman-Yee awards illustrate several broader shifts in how research institutions structure, finance, and govern artificial intelligence initiatives:

  • Demand for Verifiable Reasoning and Source Grounding: In high-stakes fields such as oncology and regulatory compliance, research teams are moving away from speculative probabilistic chatbots toward neuro-symbolic and source-checkable architectures. Investigators are embedding automated theorem provers and deterministic evidence linkages to ensure computational reproducibility across clinical datasets and published literature.
  • Equitable and Open Data Governance: Several winning projects, particularly the Prakash Lab’s diagnostic microscopy platform, treat raw training datasets as permanent global public goods. For university research administrators, supporting multi-site international collaborations requires establishing compliant data management plans that balance open science principles with patient privacy and intellectual property standards across jurisdictions.
  • Alignment with Shifting Federal and International Grant Guidelines: As public sponsors introduce formal rules on artificial intelligence disclosure and algorithmic accountability, institutional seed programs help faculty pilot compliant methodologies. Research offices tracking AI grant writing policies across NIH, NSF, and the ERC can utilize institutional incubator frameworks to test rigorous human-in-the-loop standards before faculty submit large federal proposals.
  • Translational Pilots with Regulatory and Government Entities: Rather than confining research to theoretical publications, funded teams are actively embedding pilot systems into public housing agencies, benefits offices, and regional clinics. These deployments must navigate federal oversight and emerging statutory guidelines detailed in United States AI regulatory frameworks, requiring institutional review boards and general counsel offices to formulate proactive AI compliance review procedures.

Next Steps and Future Milestones

The seven 2026 Hoffman-Yee research teams will execute their initial one-year milestones through late 2027, preparing prototypes, data pipelines, and peer-reviewed benchmark evaluations for their year-two competitive renewal presentations. Specific scheduled milestones include the public launch of the Center for AI Measurement Science (AIMS) at Stanford HAI in early 2027, which will house Sanmi Koyejo’s classroom evaluation program alongside broader AI benchmarking efforts. Meanwhile, the Kavli Institute team will integrate its multiagent astronomy platform with Rubin Observatory data streams as the observatory approaches full operational survey cadences.

Findings from the workforce upskilling TaskNet trial and legal reasoning pilots with administrative agencies are slated for preliminary release across academic workshops and policy briefs in mid-2027, providing benchmark datasets for institutional planners evaluating the economic returns of technological augmentation over labor automation.

Sources

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →