Research

We study how AI changes academic work and how universities can shape its use. These selected publications and presentations include individual scholarship by Lab members and collaborative work with colleagues across CUNY.

Publications

July 2026 · Journal article · Published

Do Artifacts Still Have Politics? Technological Determinism and Professional Agency in Academic Libraries’ AI Transformation

Stephen Zweibel

portal: Libraries and the Academy 26, no. 3 (2026): 495–512.

Compares generative AI with earlier library database automation and examines how libraries can retain professional agency when adopting systems they do not control. Proposes a critical implementation framework for making those choices.

Read article on Project MUSEFull text may require library access.

Conference presentations

June 26, 2026 · Conference presentation

Building Community-Oriented Infrastructures for AI Experimentation

Matthew K. Gold, Luke Waltzer, Zach Muhlbauer, Azucena García Gutiérrez, and Stephen Zweibel

Association for Computers and the Humanities (ACH) 2026.

Presents the Lab’s shared AI infrastructure alongside faculty development and classroom projects, including Spanish-language learning and tools for scholarly work.

View ACH presentation slides

April 21, 2026 · Conference paper set

Community, Transparency, and Tinkering for Just Futures

NARST 2026 Annual International Conference.

Four papers from the Critical AI Literacy Institute examine how teaching and faculty development can make AI systems more legible and contestable through tinkering, scholarly agency, collective curriculum design, and resistance to inevitability narratives.

Tinkering as Critical AI Literacy: Teaching Infrastructure through Breakdown & (Re)Configuration

Zach Muhlbauer

The Critical AI Literacy Institute: Asserting and Preserving Scholarly Agency in the Age of AI

Luke Waltzer

Fostering Critical AI Literacy as Collective World-Building: Curricular Models for Teaching With/About Generative AI

Laurie Hurson

Beyond the Black Box: Resisting AI Inevitability Rhetoric and Implications for Science Education

Sule Aksoy

View CALI session slides Conference program (PDF, p. 171)

Ongoing Research

Revisiting Model Organisms for Emergent Misalignment

Zach Muhlbauer

Drawing on Betley et al. (2025) and subsequent studies of emergent misalignment, this collection of Python testbeds examines how language models produce harmful or deceptive responses independent of their assigned task, using a shared evaluation harness to test and score differential outputs for misalignment.

Project and references