Starting a GenAI Community of Practice

A collaborative, low-barrier way to learn, experiment, and lead

Posted on July 7, 2025 in Blog Posts

Why AI Matters for Librarians

The year 2023 seems like a vague and distant memory. That spring, a set of librarians at our institution became aware of generative artificial intelligence. For our Computing Librarian, it was the cacophony of her field, loud with paradigm-changing promises and critiques of the new technology. For our Outreach & Teaching Librarian, the sounds of a second career. For our Social Sciences Librarian, it was a new, shiny discovery at the ACRL conference in Pittsburgh that caught the imagination. From different viewpoints, all three saw the looming specter of something. Just what it was remained unclear, but what was to come was (and in some ways still is) up for debate.

We knew we needed to act, and it was clear that something needed to be done to engage fellow colleagues about AI. Though what that “something” could be sparked so many ideas for action that any individual’s possible bandwidth might not be enough to tackle it. For all of our sakes, we were able to find each other and come up with some ideas, one of which was a community of practice. A community of practice offered a collaborative, low-barrier way to learn, experiment, and lead together in this evolving frontier.

---

🔥 Stay up-to-date with LibTech Insights by signing up for our free newsletter. Just one weekly email with our new blog posts, top tech news stories, and other bonus content. Check out some posts from our archive:

• A Reading List for Algorithmic Bias
• Why Librarians Should Apply for Tech Jobs
• Three Polling Applications to Consider for Your Instruction

🌟 Register now for our free email micro-course on AI literacy!

---

Launching the GenAI Community of Practice

Recognizing the benefits of this formal initiative, George Mason University Libraries launched a yearlong AI Community of Practice for library employees in fall 2024. A community of practice has a clearly defined, shared interest around which members engage in joint activities, discussions, and information-sharing. Members are library staff and faculty who develop shared experiences, tools, and resources, with a focus on continuous learning and improvement. The University Libraries promote communities of practice that use digital integration for the dispersed campuses and a cross-disciplinary, inclusive approach to reach as many employees as possible. We didn’t just explore what AI could do—we also asked what it should do, and how to use it responsibly in service to our patrons and profession.

The AI Community of Practice (2023–24) focused on the use of specific AI tools for work and research best practices, giving library employees the opportunity to gain experience using existing and emerging AI tools, including ChatGPT, Grammarly, Google Labs, and Bard (now Gemini). Led by a librarian experienced with using each tool, these virtual sessions met monthly for one hour and gave participants the opportunity to actively learn, experiment with, and reflect on the implications of different AI tools. These practical sessions promoted greater understanding of current and emerging applications of AI tools for our community’s day-to-day work. It also provided dedicated time and a safe space to practice, fail, and ask questions to strengthen prompt engineering skills and other foundational skills. Participants were encouraged to come to the monthly meetings to build on skills throughout the academic year, but any employee was welcome to attend a single meeting of interest. Each session averaged about four to five employees, fostering an intimate setting so the instructor could provide personalized recommendations and greater direct assistance to the participants.

After participant feedback, the AI Community of Practice (2024–25) was modified to focus on task-based themes rather than one tool per session, as the explosive growth of AI tools left many librarians wondering which was best. Each month’s topic focused on an AI-enabled task, such as image creation or text creation, with two to three tools highlighted for that application. This allowed participants to understand the pros and cons of several similar tools to find the best fit for their work.

Starting Your Own GenAI Community of Practice: A Quick Guide

Identify a Core Team: Two to three librarians with interest or experience in AI.

Define Your Purpose: Skill-building? Exploration? Policy development?

Choose a Format: Monthly virtual sessions? Hybrid workshops?

Select Tools and/or Topics: Start with accessible tools like ChatGPT, Gemini, or Canva AI, or tools readily available without a subscription

Create a Safe Learning Space: Emphasize experimentation, curiosity, and nonjudgment.

Gather Feedback and Iterate: Use surveys or informal check-ins to adapt.

Lessons Learned and Practical Tips

Our experiences with our community of practice should be prefaced with the benefit of some prior iterations. Elsewhere in our institution, we have engaged in communities of practice on a variety of topics and areas of concern and focus. This prior experience made the creation of yet another community easier, but not without challenges. Tackling a brand-new and poorly understood technology while also engaging with a “new normal” after the COVID pandemic and shifts in work/life engagement created many hurdles.

Capacity is, and will likely remain, the first and most significant barrier. For various reasons, we have real and significant limits on time and capacity from our faculty and staff. While it doesn’t sound like a big commitment to plan, organize, and execute a few sessions during a given spring or fall semester, the reality is that setting up and running a community of practice is one more thing to add to a plate full of “one more thing.” Having one or two others take on the task creates an immeasurably superior experience and offers a wider range of perspectives. Aid in planning, familiarizing, and executing in a community of practice does make (or break) the entire effort and sets the tone for others who may join.

Many other elements are going to be organization-specific. How large or small is your population? Are you keeping things within your library or opening things up to wider patron attendance and participation? Do you have physical space available with the preferred hardware to showcase web-based technology and tools? Do you have access and/or resources to obtain access to these tools? Are virtual, in-person, hybrid, or separate format sessions preferred or desirable for your population? Can people stop by for only one session, or do you want a commitment for the full length of the program? The answers are specific to your organization alone.

Looking ahead

Librarian expertise in this space is still very much being discovered and developed. Even a seemingly small bit of experience can set an individual on a pedestal. Discovering that Hugging Face, site with documentation of benchmarks and test results for various AI tools, exists (a site that won’t disappear in a month’s time and won’t date this at all) was a moment that shows the breadth of generative AI tools, the means to measure them, and even without understanding everything—just knowing that little bit more is enough. As AI continues to evolve, so must our profession. We invite other libraries to share their experiences with AI communities of practice and join the conversation.

Related Posts