Katherine Elkins.
Portrait of Katherine Elkins

Katherine Elkins

AI researcher · Kenyon College

Studying how increasingly capable AI systems reason, persuade, interpret, coordinate, and interact with people, and how to build systems that are more culturally and historically accurate.

Google Scholar CV Wikipedia

Current research

AI behavior in complex human environments

Many consequential AI questions are simultaneously technical and human: what a model does under pressure, how wording or narrative changes a judgment, when an explanation tracks behavior, and how agents coordinate or deceive.

Elkins works directly on those questions through computational experiments, behavioral testing, and interdisciplinary interpretation.

Her research spans model behavior, human judgment, narrative, agentic and multi-agent systems, behavioral evaluation, cultural memory, and governance.

Research constellation

Research constellation

Move across the map, or tap a star, to read about it.

Six research areas

AI behavior in complex human environments

All research

  1. Model behavior & human judgment

    Reasoning, ethical judgment, persuasion, deception, manipulation, calibration, and failure modes produced by framing or linguistic variation.

    Model behavior and human judgment

  2. Agents & multi-agent systems

    Agentic behavior, coordination, debate, emotional stability, and deception in systems where models interact with one another and make consequential decisions.

    Agents and multi-agent systems

  3. Narrative, cognition & emotion

    Narrative as a cognitive and emotional tool: a structure for memory and interpretation, a mechanism of persuasion, and a factor in how models and people organize judgment.

    Narrative, cognition and emotion

  4. Evaluation & robustness

    Behavioral evaluation, model comparison, ethics-based auditing, red-teaming, benchmark design, and testing that reveals vulnerabilities conventional metrics can miss.

    Evaluation and robustness

  5. Culturally & historically grounded AI

    AI systems that work with archives, cultural memory, provenance, retrieval, context, and expert interpretation, keeping knowledge tied to its context.

    Culturally and historically grounded AI

  6. AI, institutions & governance

    Standards, regulation, public-interest AI, and the institutional consequences of deploying increasingly capable models in high-stakes environments.

    AI, institutions and governance

Current project · Schmidt Sciences Humanities and AI Virtual Institute

Archival Intelligence

Building open AI tools for rescuing endangered cultural archives, and AI systems that are more culturally and historically accurate.

  • Provenance
  • Retrieval
  • Interpretation
  • Expert knowledge

Elkins is Co-PI of a collaborative project beginning in New Orleans that develops systems for preserving, organizing, retrieving, and interpreting cultural records. The goal is to work with cultural and historical materials without stripping away provenance, context, ambiguity, or expert knowledge.

Explore Archival Intelligence

Research in practice

Research leadership

  • NIST CAISI AI Consortium · 2024–present

    Measurement, evaluation, and trustworthy AI

    Elkins co-leads the five-member team representing the Modern Language Association in the NIST CAISI AI Consortium, bringing research on language, ambiguity, framing, persuasion, narrative, and human judgment into federal work on AI measurement and evaluation.

    Research context

  • Schmidt Sciences

    Archival Intelligence

    Co-Principal Investigator of a Schmidt Sciences Humanities and AI Virtual Institute project developing open AI systems for endangered cultural archives, bringing together technical research, archival expertise, cultural institutions, and research infrastructure across multiple institutions.

    archivalintelligenceai.org

  • Kenyon College · since 2016

    Human-Centered AI

    Elkins and Jon Chun founded the Human-Centered AI curriculum and lab at Kenyon in 2016, creating an interdisciplinary research model that combines technical training with domain expertise and original student research.

    AI in Higher Education Lab Mentored research

Research, education & curriculum

AI in Higher Education

Katherine Elkins’s work asks what universities should teach when students can increasingly delegate traditional academic tasks to machines. Her answer moves students upstream: from users of AI systems to researchers who evaluate, build, interpret, and question them.

Elkins and Jon Chun founded the Human-Centered AI curriculum and lab at Kenyon in 2016. The program connects technical training, disciplinary expertise, original student research, and a larger argument about what universities are for.

Explore the curriculum, lab, and research archive

Methodological foundations

Making difficult human phenomena empirically testable

Elkins’s computational research developed rigorous ways to operationalize narrative, emotion, interpretation, translation, cultural meaning, and judgment while keeping context and ambiguity in view. She now applies and extends those methods to increasingly capable models, agents, and AI systems.

Cover of The Shapes of Stories: Sentiment Analysis for Narrative (Cambridge University Press)
Cambridge University Press · 2022

The Shapes of Stories

Introduced the first rigorous methodology for narrative sentiment analysis: a reproducible framework for selecting, comparing, validating, and interpreting sentiment models across narrative texts.

Computational interpretation

Middle Reading & SentimentArcs

Middle Reading joins close interpretation to computational scale. SentimentArcs is a novel ensemble method for comparing narrative sentiment trajectories across texts.

Frontiers in Computer Science · 2024

Translation, language, and interpretation

Computational comparison makes visible what changes across translations, and what model outputs can miss when language carries cultural and interpretive context.

Selected firsts

Original methods and research designs

  1. 2016

    Human-Centered AI

    Elkins and Jon Chun founded the Human-Centered AI curriculum and lab at Kenyon.

  2. 2020

    Writer’s Turing test

    First writer’s Turing test of a large language model.

  3. 2022

    Narrative sentiment methodology

    First rigorous methodology for narrative sentiment analysis.

  4. 2023

    Explainable AI for narrative

    First application of explainable AI to narrative analysis.

  5. 2024

    Ethics-based model audit

    First ethics-based audit of moral reasoning in deployed large language models.

  6. 2024

    Comparative AI regulation

    First systematic comparison of AI regulation across the European Union, China, and the United States after passage of the EU AI Act.

Scholarly reception

Methods taken up across fields

The ethics-audit confidence-scoring method has been adopted in later LLM value evaluation. The open-model benefit–risk framework has shaped FAccT, TMLR, and governance research. Narrative methods have been extended in NLP, translation, behavioral science, persuasion research, and large-scale studies of stories.

Read the evidence

The citation record spans 15 fields and 55 subfields.

The citation record spans AI, digital humanities, literary studies, education, medicine, law, and related fields.

Speaking & exchange

Research in public

Talks and conversations address model behavior, AI governance, narrative and persuasion, cultural memory, human-centered research, higher education, and the design of AI systems grounded in cultural and institutional contexts.

Selected speaking

Contact

Research collaborations, institutes, standards work, speaking, and media inquiries

Get in touch