Schmidt Sciences Humanities and AI Virtual Institute: Archival Intelligence
Co-PI, Schmidt Sciences HAVI, 2025
Elkins is Co-PI, with Jon Chun, of Archival Intelligence: Rescuing New Orleans’ Endangered Cultural Legacy, a Schmidt Sciences Humanities and AI Virtual Institute project. The project develops open-access AI tools for preserving endangered cultural archives in under-resourced and underrepresented communities, beginning with New Orleans collections tied to Creole, Cajun, and jazz heritage.
The project was selected as one of twenty-three teams worldwide working at the intersection of AI and the humanities and provides up to $330,000 over eighteen months to build AI-powered tools for cultural preservation.
Project site → Schmidt Sciences → News →
Notre Dame–IBM Technology Ethics Lab Grant
PI, 2024
Elkins received a Notre Dame–IBM Technology Ethics Lab grant for “How Well Can GenAI Predict Human Behavior? Auditing State-of-the-Art Large Language Models for Fairness, Accuracy, Transparency, and Explainability (FATE).” The project, with Jon Chun and Notre Dame faculty collaborator Yong Suk Lee, examines how LLMs reason in high-stakes decision-making over humans.
Using recidivism prediction as a test case, the project compares human experts, statistical machine-learning models, and LLMs under a FATE framework: Fairness, Accuracy, Transparency, and Explainability. It asks whether generative AI systems can provide responsible predictive or judicial decision support in contexts where errors can affect human lives.
Notre Dame → Notre Dame →
U.S. AI Safety Institute Consortium / CAISI
Co-leading the MLA-sponsored team, 2024–present
Elkins co-leads the team representing the Modern Language Association at the U.S. AI Safety Institute Consortium, now housed within CAISI. The MLA-sponsored team brings expertise in language, writing, interpretation, ethics, and humanistic inquiry to evaluations of model behavior in complex linguistic and ethical scenarios.
The team’s ethics-based audit results were presented on its behalf during the opening keynote at the consortium’s first plenary at the University of Maryland. This placed humanistic model-evaluation work inside one of the central U.S. standards conversations about trustworthy AI.
NIST → MLA → MLA →
ICML 2024 Oral Presentation
International Conference on Machine Learning, 2024
Elkins was part of the international author team for “Position: Near to Mid-term Risks and Opportunities of Open-Source Generative AI,” selected as an oral presentation at ICML 2024. The paper operationalizes an AI openness taxonomy across forty large language models and analyzes risks, benefits, and governance needs associated with open-source and open-weight generative AI.
The author group brought together researchers from Oxford’s Foerster Lab for AI Research and Torr Vision Group, UC Berkeley, Notre Dame, Kenyon, Berklee, the Luxembourg Institute of Science and Technology, the University of Luxembourg, the University of Bath, ITESM, the National University Philippines, and other international institutions.
PMLR → arXiv →