AI + OER in Practice: Building a Community-Centered Model for Responsible Innovation

Judith Thomas; Peter Musser; Megan Simmons; and Bethany Mickel

Open educational resources (OER) have long been positioned as a means of expanding access to course materials, enabling pedagogical innovation, and reducing costs for students. However, the development and adaptation of OER remains labor intensive and is often under resourced. Faculty and instructional teams frequently face significant time constraints, limited funding, and gaps in ancillary materials needed to support effective teaching.

Within this context, the rapid emergence of generative artificial intelligence (Gen AI) presents both compelling opportunities and complex challenges. Gen AI can accelerate content creation, support accessibility improvements, and assist with adaptation and localization. Yet it also raises critical questions related to quality, bias, authorship, licensing, privacy, and the role of human expertise and experience in educational work.

This project emerged at the intersection of these tensions. Rather than focusing on a single tool or technical solution, it asks a broader question: how can institutions thoughtfully engage with AI in ways that are aligned with the values of open education?

To address this question, the initiative centers on community infrastructure and shared learning. It positions AI not as a replacement for human labor, but as a tool to be explored collaboratively, critically, and transparently within a broader ecosystem of open educational practice.

Project Overview and Partnership Model

This initiative was developed as a pilot project through a partnership between the University of Virginia (UVA) Library and the Institute for the Study of Knowledge Management in Education (ISKME), with funding from the Institute of Museum and Library Services (IMLS).

From the outset, the project was designed to address two interconnected needs: creating shaed infrastructure for documenting and disseminating emerging practices at the intersection of AI and OER, and fostering sustained community engagement to explore the opportunities and risks of Gen AI in educational contexts. Rather than developing a standalone tool or platform, the project adopts a community centered model built around two complementary components: An AI & OER Community Hub hosted on OER Commons and an AI & OER Fellowship Program that supports collaborative exploration, knowledge creation, and the sharing of emerging practices.

These components are intended to support each other’s development: the Fellowship Program generates experimentation, dialogue, and resources, while the Hub provides a public-facing space for sharing, curation, and ongoing community engagement.

The design has four guiding principles:

  • foreground open education values
  • prioritize community over technology
  • use AI according to agreed-upon standards
  • support iterative, practice-based learning

The AI & OER Community Hub

The AI & OER Community Hub is a digital space within OER Commons designed to support resource sharing, collaboration, and knowledge exchange across institutions and sectors. The purpose of the pilot project was to lay the groundwork for future development of the hub.

At its core, the Hub is a digital library, offering curated collections of materials related to AI and OER. It also aims to serve as a community space by creating opportunities for discussion, contribution, and collaboration among educators. The inclusion of an events calendar connects users with a community-populated listing of webinars, conferences, and professional development opportunities related to AI and education. Hosting the Hub on OER Commons allows the Hub to evolve with new developments and changes in the education landscape.

The Hub’s collections are grouped into eight broad themes based on expected user’s needs (e.g., education level or topic of interest), within which they can find relevant collections. For example, the themes for K-12 and higher education each contain collections corresponding to grade levels as well as instructional design, while the ethics and law theme contains collections concerning copyright, misinformation, environmental impact, and more.

Approach to Curation

New tools, frameworks, and discussions on the topic emerge on a daily basis. This becomes another barrier for many educators, OER practitioners, and librarians who already spend significant time searching for resources that meet their diverse needs and standards of quality.

To address this, the project adopts an approach that treats curation as a core activity rather than a secondary task. OER Commons’ more than 10,000 daily users regularly find and submit items to the collection. ISKME’s staff librarians review all submitted materials according to the library’s collection policy, enriching approved resources’ metadata to improve searchability while rejecting materials when appropriate. This work has become particularly important since the emergence of AI tools has also led to a high volume of poor quality resources created by tools with little to no human review.

This approach remains grounded in the values of open education, including a commitment to accessibility, equity, and reciprocity.

The AI & OER Fellowship Program

The AI & OER Fellowship Program was a year-long, cohort-based initiative that brought together practitioners from K–12 and higher education to explore the intersection of AI and OER.

Participants were selected through a competitive process, with attention to diversity of roles, institution types, disciplinary backgrounds, and geographic locations. The resulting cohort included educators, librarians, instructional designers, and technologists, each bringing perspectives and expertise.

The Fellowship Program combined regular engagement with varied means of participation. It featured twelve monthly synchronous sessions that provided dedicated space for shared learning, alongside ongoing discussions that allowed conversations to develop and evolve over time. Central to the Fellowship was exploration, with participants engaging directly with a variety of AI tools and use cases, experimenting and reflecting on their potential applications in educational contexts. This hands-on engagement fostered practical understanding while also surfacing important questions about effectiveness, ethics, and accessibility.

Curation was another key focus as Fellows identified, evaluated, and organized resources for inclusion in the Hub, collaborating with ISKME’s librarians to curate a collection of resources that met their quality criteria around openness, quality, and AI relevance. This curatorial process emphasized critical reflection, ensuring that materials were relevant and actionable for educators working with open educational resources. In addition to curation, some participants developed openly licensed instructional materials and models that could be adapted and shared widely, extending the impact of their work, through resources such as AI by Design: A Framework for Educators Building AI-Powered Learning Tools (Moy, 2026), AI PD Sessions for Administrators (Cobbs, 2026), Bloom’s StAIrcase (Meskell & Baker, 2026), Trial of Socrates with CustomGPT Socrates Bot Lesson (Sylvia, 2025), and Exploring the Dynamics of Artificial Intelligence in Higher Education (Frazier, 2026).

Rather than applying a fixed rubric to OER evaluation, the project took a more flexible approach, recognizing that standards for evaluating AI-related resources were still evolving. Fellows engaged in ongoing discussions about what constituted a high-quality AI-related resource within an open context, drawing on their experiences and perspectives to shape this understanding. In doing so, they considered factors including the relevance of a resource to OER workflows and practices, the clarity and reliability of the content, and its accessibility and usability for different audiences. They also attended to the ethical implications of AI use, recognizing that questions of responsibility and impact were central to evaluating these tools and materials.

Finally, the Fellowship emphasized community building, fostering a sustained network of practice that supported collaboration, dialogue, and knowledge exchange. Through collaborative curation, resource development, and shared exploration, Fellows cultivated connections intended to endure beyond the structured sessions.

Session topics, discussions, and practice activities were intentionally scaffolded, beginning with foundational priorities to guide this work, such as articulating ethical considerations, exploring Gen AI tools and best practices, and identifying quality criteria for resource evaluation. Sessions progressed to areas such as designing for accessibility, developing research approaches, incorporating open pedagogy, and developing leadership, advocacy and outreach communication skills.

A key feature of the Fellowship was its emphasis on reflection and dialogue. Participants were encouraged to share their experiences, question assumptions, and engage in critical discussion. This created a space where uncertainty was acknowledged and explored, rather than minimized.

The Fellowship proved particularly valuable in addressing a common challenge: many practitioners were experimenting with AI in isolation within their own institutions. By bringing participants together, the program emphasized shared learning and decision-making informed by the perspectives of a diverse community.

Lessons Learned

Several key lessons emerged, highlighting both challenges and opportunities in AI-supported open education.

First, community must precede infrastructure. The effectiveness of the Hub depends less on its technical features than on engagement, trust, and shared purpose. Sustained participation depends on creating opportunities for people to work and learn together, rather than in isolation.

Second, disagreement is useful. AI in open education is still evolving, and participants brought different perspectives, including skepticism.  Several fellows indicated that their attitudes towards AI changed over the course of the year, as the impacts of AI on research, teaching, and scholarship became more marked.  Making space for these views led to more careful thinking and stronger outcomes.

Third, curation requires deliberate effort. It is not a passive activity, but one that depends on time, expertise, and clear criteria. Providing structure and support helped participants do this work well and maintain the quality of the Hub.

Finally, values must be put into practice. Principles such as transparency, equity, and accessibility need to guide day-to-day decisions—how resources are selected, how tools are used, and how questions of authorship and access are handled.

Assessment

Our approach to assessment reflected the project’s exploratory, community-based nature, combining quantitative indicators with qualitative inquiry to capture activity and impact.

Measures included participation indicators such as numbers of Hub users and contributors, volume and diversity of resources, and fellowship participation and retention. These were complemented by semi-structured interviews examining how participants’ goals aligned with their experiences, how they engaged with peers, and what supported sustained involvement.

Interview findings indicated that the Fellowship deepened participants’ understanding of AI and OER, moving many toward more applied engagement. Fellows emphasized the value of cohort-based interaction and dedicated time for discussion, and many reported increased confidence and visibility as leaders, extending their work beyond the program, sharing practices, resources, and ideas within their institutions and professional networks. Feedback also highlighted participants’ desire for structured opportunities for collaboration.

Participants also noted areas where the program could be strengthened. Given the diverse mix of K12 and higher education professionals, a more tailored focus–perhaps through differentiated tracks–might better have met varying participant needs. Furthermore, evaluation data revealed a desire for structured collaborative opportunities, including for the joint creation of OER.

Assessment also considered contributions to the Hub, including the number and types of resources added, as well as instructional materials, models, and evaluation approaches developed by Fellows. Interviews and reflective prompts illuminated how participants approached curation, assessed resource quality, and refined their thinking about ethical and effective uses of AI. Fellows noted both the value of a shared space for resource discovery and the challenges of identifying high-quality materials that integrate AI and OER in specific contexts.

Usability testing provided additional insight. Fellows completed common tasks—such as uploading resources or navigating features—while describing their process. These sessions identified barriers and uneven awareness of Hub functionality.

If extended beyond the pilot phase, assessment will focus on longer-term outcomes, including changes in teaching practices, patterns of reuse and adaptation, and community sustainability.

Implications and Replicability

This project offers a model that others can adapt as tools and technologies continue to change. One key insight is that the work should not be tied to any specific platform, but instead focus on how different elements of a program work together. Pairing a public Hub with a cohort-based Fellowship created a balance between documenting resources and sharing ideas.

Several aspects of this approach may be useful elsewhere. Treating curation as a core activity helps address the challenge of navigating a rapidly expanding body of AI-related information. The cohort model provides a structured space for participants to learn from one another, test ideas, and refine their thinking over time.

Grounding the work in open education values shaped how participants evaluated resources, approached authorship, and considered issues such as access and ethics.

More broadly, the project demonstrates an alternative to both uncritical adoption of AI and outright resistance by emphasizing shared inquiry, visible practice, and ongoing engagement.

While the structure of a hub or fellowship may vary, the underlying approach—linking shared infrastructure with community, supporting curation and creation, and allowing evaluation practices to develop alongside the work—can translate across institutions. Adaptation to local needs and constraints will be more important than direct replication.

Finally, the project highlights that the main challenges at the intersection of AI and OER are not technical, but social and organizational: how people make sense of new tools, share knowledge, and develop common practices.

References

AI & OER Community Hub. OER Commons. https://oercommons.org/hubs/ai-in-oer

Cobbs, T. (2026). Quick AI PD Sessions: Administrator Templates. OER Commons. https://oercommons.org/courseware/lesson/142544

Frazier, A.T. (2025). Exploring the Dynamics of Artificial Intelligence in Higher Education. OER Commons. https://oercommons.org/courses/exploring-the-dynamics-of-artificial-intelligence-in-higher-education

Meskell B. & Baker N. (2026). Bloom’s StAIrcase. OER Commons. https://oercommons.org/courses/bloom-s-staircase

Moy, M. (2026). AI by Design: A Framework for Educators Building AI-Powered Learning Tools. OER Commons. https://oercommons.org/courseware/lesson/142573

Sylvia, J. J. IV (2025). Lesson: Trial of Socrates with CustomGPT Socrates Bot. OER Commons. https://oercommons.org/courseware/lesson/130581

License

Icon for the Creative Commons Attribution 4.0 International License

AI + OER in Practice: Building a Community-Centered Model for Responsible Innovation Copyright © 2026 by Judith Thomas; Peter Musser; Megan Simmons; and Bethany Mickel is licensed under a Creative Commons Attribution 4.0 International License, except where otherwise noted.