Equity and AI: Paper Session Summary
In an effort to share ideas beyond the limited time we have during the annual conference, the Community Building Committee asked Graduate Students attending TPRC54 to take notes of paper sessions to be made available to the community. The following is a summary of the paper session entitled “Equity and AI” that was presented on Friday, September 25, 2026 at 4:00pm. The session was moderated by Christopher Mohawe, University of Illinois Chicago. The following is an edited summary based on the notes provided by Ebubechukwu Uba, University of Illinois Urbana-Champaign.
Paper One: Building Equitable AI Ecosystems: Institutional Capacity, Partnerships, and Policy Implications at a Minority-Serving Institution
Presenter: Christopher Lawson, North Carolina Central University
Additional Authors: Shantel Reddick, North Carolina Central University; Clinton Griffin, North Carolina Central University; Siobahn Day Grady, North Carolina Central University
The presentation examined how artificial intelligence is transforming education and the workforce and emphasized the importance of ensuring that marginalized communities are included in conversations about AI. Minority-serving institutions (MSIs) are particularly affected by the rapid development of AI, especially because of challenges such as limited federal funding and the lack of national AI literacy standards.
North Carolina Central University (NCCU) responded through its Institute for AI and Emerging Research (IAIER), using the principle “AI is for everyone.” The initiative focused on building institutional capacity in AI development, workforce development, and community development. The research examined institutional documents to explore how AI has shaped literacy, how partnerships build ecosystem capacity, and what policy lessons can be learned for MSIs.
Key findings included:
Structured programming can drive AI literacy.
Faculty development contributes to institutional capacity building.
Students benefit from clearly defined AI pathways.
Partnerships can serve as important policy instruments.
MSIs can play meaningful roles in public-sector AI evaluation.
The presentation also addressed concerns about AI automating routine jobs, including entry-level technical work, basic data cleaning, and routine coding. Infrastructure limitations remain a challenge, demonstrating the need for policy solutions at the institutional, state, and federal levels.
Paper Two: The Financial Structure of AI Companies
Presenter: Mihir Kshirsagar, Princeton University
Additional Author: Felix Chen, Princeton University
This presentation focused on the financial structure of AI and big technology companies. The presenters asked important questions about who will benefit from massive AI spending and who is responsible for ensuring that this spending is used effectively.
The AI market is highly concentrated, with major companies such as Nvidia, Amazon, and Microsoft playing significant roles across the industry. The presenters discussed concerns involving market concentration, vertical integration, and the pressure for major companies to generate returns on their investments, potentially at the expense of users.
One potential path forward is maintaining viable open-weight AI models, which could provide competition against expensive hardware and computing resources.
Policy recommendations included:
Keeping open weights viable.
Funding public computing resources.
Requiring portability.
Keeping distribution channels open.
Separating AI models from cloud providers.
Lightning Talk: Privacy, AI, and Libraries
Presenter: Laila Moustafa, University of Illinois
This lightning talk examined the relationship between AI, privacy, and academic libraries. The central question was whether libraries' traditional commitment to privacy can survive the increasing use of AI.
The study used a mixed-methods approach, including eight in-depth interviews with academic library professionals nationwide and a survey of 65 leaders based on themes identified through the interviews.
The research identified privacy concerns related to how AI evaluates people's data, including digital footprints and digital identities. The presentation argued for AI-specific privacy policies that place responsibility on institutions.
Important recommendations included:
Providing clear notice when AI tools are being used.
Obtaining appropriate consent.
Increasing transparency around AI systems.
Including AI considerations in campus decision-making committees.
Session One Reflection
“The first paper was great to see that NCCU was the first HBCU to have an AI and Emerging Research Center, and they are doing great work to create great AI development for the workforce and for future generations of students. Learned about the financial bubble in Big Tech companies.
Do we want to keep things more traditional in libraries, or are there alternatives that can be explored if a user is skeptical enough to want to opt out of having their data assessed by AI? – This was answered in the Lightning Talk, where the speaker says she is also wondering the same if she is trying to isolate people from technology development, but she thinks it might be a matter of choice of words, and what would solve this is transparency. Transparency is able to solve issues of privacy.
Great presentations.”