501(c)(3) Nonprofit Research Organization

Human-Centered AI Lab: AI research and design that prioritizes human thriving.

We study how AI systems reason, persuade, inherit cultural assumptions, and act where human judgment matters.

00 — Premise

The questions that animate our work are old ones: what kind of minds we are building, whose traditions they carry, and what they mean for human flourishing.

As an independent 501(c)(3) nonprofit, the Human-Centered AI Lab connects technical evaluation with expertise in language, culture, policy, and history.

Our work tests models, compares governance systems, and develops methods for archives and cultural data.

Research downloads: 100,000+
Countries reached: 198
Institutions: 4,000+
Research projects: 400+

01 — The Questions

Three ways in.

AI Safety & Governance Research

We test how models handle rules, negation, persuasion, and ethical judgment; compare governance regimes; and study oversight for autonomous systems. The Wisdom Test asks which religious and philosophical traditions frontier models reproduce or leave out.
Interdisciplinary Collaboration

We form focused teams around questions that require technical researchers, humanists, social scientists, policy scholars, and cultural institutions to work together.
Open Tools & Public Scholarship

We publish methods, code, and research that others can inspect and extend, from model-evaluation protocols to tools for narrative analysis and endangered archives.

02 — Current Work

NIST · Center for AI Standards and Innovation MLA Principal Investigators

NIST CAISI

Katherine Elkins and Jon Chun lead the Modern Language Association’s consortium team, contributing humanities and language expertise to federal AI standards.

Agent security. How agents interpret natural-language instructions rather than simply executing them.
Interpretive tractability. Whether human overseers can meaningfully evaluate an agent’s actions and reported reasoning at the pace agents operate.
Benchmarks. Evaluation must account for the languages, cultural assumptions, and prompt formulations through which capabilities are measured.

Syntactic fragility: small structural changes in otherwise equivalent prompts can reverse a model’s recommendation.

NIST CAISI MLA Newsletter, Summer 2026

Schmidt Sciences · Humanities and AI Virtual Institute One of 23 teams selected worldwide

Archival Intelligence

AI methods to preserve, restore, and analyze endangered multilingual and multimodal archives without stripping away provenance, cultural context, or community authority.

Archival photograph: two students reading periodicals at a library magazine rack. — FSA/OWI photograph · Library of Congress

Early work includes poorly digitized New Orleans cultural materials. The award is administered through Kenyon College.

archivalintelligenceai.org NPR / WOSU coverage

Featured initiative A successor to the Turing Test

The Wisdom Test

The question is no longer whether a machine can pass for human, but what kind of mind it presents.

The Wisdom Test probes views of selfhood, obligation, the sacred, suffering, and fate to reveal which philosophical and religious traditions frontier models reproduce or omit.

4 of 8 axes piloted · planned: an open multilingual archive and reusable evaluation methods

Read about the Wisdom Test

03 — Research index

Eight lines of inquiry.

We evaluate models, compare governance systems, reconstruct damaged archives, and study how machines handle language, culture, and human judgment.

AI Systems
03.1 LLM Evaluation & Red-TeamingTests of syntactic fragility, negation, prohibition, persuasion, and reported reasoning, contributing language-centered methods to the MLA team’s research in NIST CAISI. Evaluation
03.2 Multi-Agent Behavioral SimulationMore than 90 model and reasoning combinations tested on high-stakes decisions, including judicial recidivism prediction, with support from the Notre Dame–IBM Technology Ethics Lab. Behavior
03.3 The Wisdom TestA successor to the Turing Test: which philosophical and religious traditions do frontier models reproduce or omit? Four of eight axes piloted. Featured initiative
Governance & Society
03.4 Comparative AI PolicySystematic comparison of AI regulation in the European Union, China, and the United States, with work on open models, behavioral evaluation, and institutional accountability. Regulation
03.5 Provenance & Cultural CommonsRights-aware training, provenance infrastructure, and community control of cultural data, developed through the AI, IP & Culture Repository co-design process. Cultural data
Culture & Interpretation
03.6 Archival IntelligenceAI methods to preserve, restore, and analyze endangered multilingual and multimodal archives. Supported by Schmidt Sciences HAVI. Cultural preservation
03.7 SentimentArcs & Multimodal AnalysisOpen methods for tracing emotional patterns across full-length texts and film. Narrative
03.8 Human-Centered AIThe 2016 curriculum and 2023 research framework joined computational practice to ethics, governance, and interpretation. Foundational work

03 — Selected work

On the shelf.

Complete, up-to-date lists live on the founders’ scholarly profiles: Katherine Elkins (Google Scholar) and Jon Chun.

Book · 2022
Elkins, K. The Shapes of Stories: Sentiment Analysis for Narrative. Cambridge University Press.
Book · 2022
Elkins, K., ed. Proust’s In Search of Lost Time: Philosophical Perspectives. Oxford University Press.
Article · 2020
Elkins, K., & Chun, J. “Can GPT-3 Pass a Writer’s Turing Test?” Journal of Cultural Analytics.
Article · 2023
Chun, J., & Elkins, K. “The Crisis of Artificial Intelligence: A New Digital Humanities Curriculum for Human-Centered AI.” International Journal of Humanities and Arts Computing.
Policy · Summer 2026
Elkins, K. “The MLA’s Work on Federal AI Standards.” MLA Newsletter, vol. 58, no. 2, pp. 1–2.
Chapter · Forthcoming
Elkins, K. “The Dual Edge: AI’s Impact on Cultural Production and Preservation.” Artificial Intelligence and Culture: A Global Perspective from Cairo. UNESCO Regional Office for Egypt and Sudan.
Under review · AIES 2026
Two papers on cross-cultural ethical auditing of frontier models. Titles withheld pending double-blind review.

Grants & affiliations

NIST · Center for AI Standards and Innovation
Schmidt Sciences · Humanities and AI Virtual Institute
Notre Dame–IBM Technology Ethics Lab
UNESCO · AI, Culture & Heritage
OpenAI Higher Education Forum
Bloomberg · AI Strategy

04 — The Lab

The people.

A small leadership team and an independent board connect model evaluation, public administration, product development, history, and policy.

Co-Founder & Co-Director

Katherine Elkins

AI safety researcher and Professor of Humanities at Kenyon College. Principal Investigator for the Modern Language Association team at NIST CAISI and of Archival Intelligence through Schmidt Sciences HAVI.

Author of The Shapes of Stories (Cambridge UP, 2022) and Proust’s In Search of Lost Time: Philosophical Perspectives (OUP, 2022). Ph.D., UC Berkeley.

Website Research Kenyon profile

Co-Founder & Director

Jon Chun

AI research scientist and co-creator of the first human-centered AI curriculum. Created SentimentArcs, an open-source toolkit for diachronic sentiment analysis; ICML 2024 oral presentation (top 2%). Co-PI for the MLA team at NIST CAISI and of Archival Intelligence.

Co-founded SafeWeb ($26M acquisition by Symantec; first In-Q-Tel security investment). UC Berkeley EECS, UT Austin MS. Two US patents.

Website Research Kenyon profile

Board of Directors

Fredrika Pfeiffer Public-sector operator with experience in contracts, procurement, compliance, and project delivery; studied Sociology and Data Analytics at Kenyon College.
Raul Romero Applied-AI builder and entrepreneur with experience at eBay and Kite ML; pursuing a master’s in design engineering at Harvard.
Jennifer Siegel Historian of intelligence, diplomacy, and statecraft; Bruce R. Kuniholm Distinguished Professor at Duke’s Sanford School of Public Policy.

2016Elkins and Chun develop a human-centered AI curriculum in the Integrated Program in Humane Studies at Kenyon College.
2023The Lab is organized.
Mar 2026Incorporated in Ohio as Human-Centered AI Lab, Inc.
Jun 2026IRS recognition as a 501(c)(3) public charity.

05 — Dispatches

From the Lab, and about it.

Announcements from the Lab and the people connected to it, followed by external media coverage.

Lab news

In the press

06 — FAQ

Frequently asked.

Common questions about the Human-Centered AI Lab, our research, and our initiatives.

01What is the Human-Centered AI Lab?

An independent research nonprofit that evaluates AI systems and studies their cultural, social, and institutional consequences.

02Is it a nonprofit?

Yes. Human-Centered AI Lab, Inc. is an Ohio nonprofit and IRS-recognized 501(c)(3) public charity. Donations are tax-deductible as permitted by law. EIN: 41-4616677.

03Is the Lab part of Kenyon College?

No. The Lab is independently incorporated and governed. Its founders teach at Kenyon, and Kenyon collaborates on or administers some projects, but the Lab is not a Kenyon center or academic program.

04What is Archival Intelligence?

Archival Intelligence is a distinct research initiative that develops AI methods for endangered cultural archives, with attention to provenance, multilingual materials, and community authority over cultural data.

05How is the Lab related to Humane Studies?

The Integrated Program in Humane Studies is a Kenyon College academic program where much of the founders’ educational work began. The Lab is a separate nonprofit research organization.

06What kinds of collaborations does the Lab undertake?

Projects include model evaluation and standards, comparative AI policy, behavioral simulation, cultural-data governance, archival AI, and public scholarship. Partners may include researchers, standards bodies, funders, cultural institutions, and technical teams.

07Can researchers or institutions partner with the Lab?

Yes. The Lab can join grant proposals, convene research teams, serve as an awardee or subawardee, and support shared research infrastructure. Start with the contact section.

08Where can I find publications and code?

See Publications for selected work and Research for project links. The footer links to GitHub repositories and the founders’ complete scholarly profiles.

09Can I support the Lab?

Yes. The Lab welcomes tax-deductible gifts and conversations with prospective funders. Write to info@humancenteredailab.org for donation instructions or due-diligence materials.

07 — Contact

Collaborate.

We welcome specific research proposals: model evaluations, policy comparisons, cultural-data projects, archival tools, and grant collaborations that need both technical and domain expertise.

info@humancenteredailab.org

The Lab can serve as prime awardee, subawardee, or research partner and maintains dedicated banking and grant administration.

Key facts

Legal name: Human-Centered AI Lab, Inc.
Status: Independent 501(c)(3) public charity; IRS determination received June 2026.
EIN: 41-4616677
Incorporated: Ohio, March 2026 (organized 2023)
Co-founders: Katherine Elkins and Jon Chun
Board: Fredrika Pfeiffer, Raul Romero, Jennifer Siegel
Key affiliations: Schmidt Sciences HAVI, NIST CAISI (via the Modern Language Association), Notre Dame–IBM Tech Ethics Lab
Official profiles: X · Wikidata · ProPublica Nonprofit Explorer · Candid · LinkedIn · GitHub