Envisioning AI’s Role in Libraries
Perspectives on Innovation, Equity, and Responsibility Across Career Stages
© 2026 Russell Michalak, Trevor A. Dawes, and Ava Wallace
This article grows out of a professional dialogue among three librarians working at different career stages: Ava Wallace, an early career archival intern and LIS student; Russell Michalak, a mid-career library director at a Hispanic-Serving Institution; and Trevor A. Dawes, an established university librarian and national leader. Their exchange reflects a shared commitment to examining both the opportunities and the challenges that artificial intelligence (AI) brings to libraries.
Rather than advancing a single argument, the contributors present their perspectives in conversation, showing how AI is reshaping discovery, access, and professional practice while also raising pressing ethical and equity concerns. The discussion is framed around three interlocking themes:
- Innovation: The technical pathways that AI opens for cataloging, metadata enrichment, and discovery
- Equity: The licensing strategies, policy frameworks, and collaborative infrastructures needed to ensure that access is just and inclusive
- Responsibility: The professional and ethical obligations to preserve human judgment, sustain workforce pathways, and prepare patrons to navigate AI critically
Ava underscores these stakes early by noting that AI can unlock powerful technical possibilities in archives and libraries. She points to “incredible search capabilities that can elucidate connections and locate hyper-specific data points,” enabling research that would have been impossible when records “didn’t exist in a viewable public form.” Seen alongside Trevor’s vision of making hidden collections visible and Russell’s emphasis on workforce preparation, her observation highlights how different career stages converge on the same recognition: AI’s potential to transform discovery is real, but its impacts depend on how librarians choose to shape and guide its use.
What Automation Promises—and What It Risks
Trevor sees AI’s most immediate promise in its ability to make hidden collections visible. “Much of our special collections, local history materials, and non-English resources remain underutilized due to poor indexing or the need for specialized knowledge to access them,” he notes. For him, tools that generate metadata, transcribe handwritten texts, and translate rare materials are not simply about efficiency but also about democratizing access. In his vision, a family historian might discover letters long buried in unprocessed boxes, or a student might finally reach sources in a language they do not speak.
Ava, while sharing this sense of possibility, urges caution. She acknowledges the promise of AI for creating new opportunities for user engagement and improving library workflows but also warns of the risks that come with widespread automation. “In the worst case, increased automation depersonalizes library services, alienates users from developing relationships with their librarians and library staff, and current tensions around job security, privacy, and impacts on mental health will continue to escalate,” she reflects. For her, the future of the field rests on balancing technical opportunities with a human-centered approach.
Russell bridges these perspectives by situating automation within the context of student learning. He shares his experiences with JSTOR’s Seeklight, where generative AI helped students surface connections across disciplines and approach scholarship from new angles. He also describes student-led digital archiving projects, in which AI-driven metadata enrichment allowed students to curate institutional memory while gaining hands-on experience with emerging technologies. In Russell’s view, automation is valuable not just for accelerating workflows but for creating opportunities that prepare students for an AI-driven job market. At the same time, he stresses that these initiatives succeed only when supported by clear institutional policies—licensing agreements that ensure equitable access, training frameworks that build ethical awareness, and guidelines that protect privacy and transparency. For him, automation is not a stand-alone solution but rather part of a broader policy commitment to align technical innovation with educational equity.
Together, this exchange frames automation and discovery as both possibilities and dilemmas. Trevor highlights its capacity for access, Ava underscores its risks for alienating workers and patrons, and Russell demonstrates how it can be harnessed for workforce development. Their dialogue makes clear that the challenge is not whether to automate but how to do so in ways that preserve human connection and open pathways for future professionals. But ensuring that these technical gains are meaningful also depends on who gets access to the tools in the first place, which shifts the conversation toward equity and licensing.
Who Gets Access, and on What Terms?
Equity quickly surfaces as a shared concern. Ava begins by voicing skepticism about the way AI is often framed as a solution to structural problems. “Given that public services are in a constant state of funding crisis, I don’t blame people for seeking band-aids,” she explains, “but that is how we should discuss them—as emergency measures—while we advocate for the resources communities truly need.” At the same time, she admits that democratized access, even when imperfect, can prevent inequities from widening. The California State University (CSU) system’s decision to provide every student with a premium ChatGPT account, for instance, struck her as “dismaying” in principle but preferable to a future where only wealthier students could afford such tools.1
Trevor picks up on this tension, shifting the conversation to the national stage. He argues that isolated fixes like CSU’s decision underscore the need for coordinated frameworks. “Leadership organizations need to move beyond issuing guidelines to creating practical, scalable frameworks that work for all types of institutions,” he insists. Cooperative purchasing agreements, shared infrastructure, and enforceable privacy standards could ensure that a rural public library has access to the same foundational AI tools as a well-funded research university. For Trevor, equity cannot be left to chance or piecemeal solutions—it must be designed into the profession’s infrastructure.
Russell brings the discussion back to the local level, where these broader debates play out in daily decisions about licensing. At Goldey-Beacom College, a Hispanic-serving institution where many students are first generation and balancing work with school, he emphasizes that “equity is a lived necessity.” By securing campuswide licenses, he ensures that no student is excluded from AI literacy because of their ability—or inability—to pay. His approach mirrors Trevor’s call for scalable frameworks and resonates with Ava’s concerns about systemic inequities, but it also shows how immediate policy choices at the institutional level can make abstract commitments to equity tangible.
Together, their exchange illustrates that equitable access to AI requires action at multiple scales. Ava reminds us that stopgap measures cannot substitute for long-term investment, Trevor presses for systemic solutions, and Russell demonstrates how campus policies can embody these principles in practice. Yet equitable access alone is not enough—students and patrons must also know how to use AI responsibly, which brings the conversation to questions of literacy and human judgment.
Teaching AI—or Teaching Against It?
As the conversation shifts from access to literacy, Trevor raises a cautionary note. He argues that the real challenge of AI is not just its technical power but also its potential to blur the line between credible synthesis and superficial mimicry. “Librarians will need to become AI translators,” he suggests, guiding patrons through outputs that often appear authoritative but can be misleading. For him, literacy is less about learning to operate tools and more about modeling habits of inquiry: asking where information comes from, why it looks convincing, and what might be missing.
Ava agrees, but she presses for an even broader definition. “Literacy does not only mean how to use something; it also means how to think critically about it, how to assess its accuracy, and how to determine when it is or is not an appropriate tool to turn to.” She insists that AI literacy must be framed as discernment—an ability to resist the coercive normalization of AI in everyday life. “In terms of implementing AI in library settings, I see opportunity and risk as often two sides of the same coin—the capabilities enabled by AI are neither inherently progressive nor inherently harmful; what matters is how they are being utilized, for what purposes, and with what effects,” she says. “As library professionals, our role is to steer the ship—both within our own organizations and at the societal level—as thought leaders in navigating technological shifts, evolving knowledge communities, and changing social norms of information seeking and sharing.” Russell connects these ideas directly to student learning and workforce preparation. At Goldey-Beacom College, he incorporates AI into student employment and training through digital archiving projects. By teaching student workers to use AI tools for metadata enrichment and institutional memory, he prepares them not only to support library workflows but also to enter a job market where AI fluency is increasingly expected. Yet, echoing both Trevor and Ava, he stresses that technical proficiency alone is not enough. Ethical awareness and critical questioning must be built into the training. In his words, AI literacy is inseparable from preparing “informed, critically engaged citizens.”
Across their dialogue, a common theme emerges: Literacy must preserve human judgment rather than diminish it. Trevor emphasizes translation, Ava insists on discernment and resisting AI hype, and Russell highlights education and empowerment. And because these challenges cannot be addressed in isolation, the next step is to consider how collaboration—within institutions, across generations, and across the profession—can sustain ethical and responsible practice.
Can We Navigate AI Alone? Why Collaboration Matters
When the conversation turns to collaboration, Trevor starts by widening the lens. For him, the question is how libraries can build structures that no single institution could achieve alone. He points to organizations like ARL and ACRL, arguing that their role must go beyond issuing statements of principle. “We need cooperative purchasing agreements, enforceable privacy standards, and shared AI models,” he explains. Without these collective frameworks, he worries that resource-rich universities will accelerate ahead while smaller institutions are left behind.
Ava responds by emphasizing that collaboration is not just structural but cultural. She cautions against allowing debates over AI adoption to fracture the profession, warning that “We need to stay grounded in what we do share, which is a commitment to the well-being of our LIS workers and library patrons. While we may not always agree, we cannot let divisions and disagreements about how to engage with AI foreclose critical conversations and collaboration.” She thinks librarians’ primary role is to bring our professional expertise in knowledge production, promoting literacy and navigating informational terrains to help users face the increasing implementation of AI from a place of understanding, empowerment, demystification, and awareness. For Ava, strong community engagement, intergenerational collaboration, and a willingness to experiment are all necessary so that AI is not treated as a mandate from above but rather as a collective process of negotiation and reflection.
Russell picks up on both threads and brings them into the day-to-day realities of library work. At Goldey-Beacom College, he has woven collaboration into student employment, where digital archiving projects become opportunities for mentorship. “When students learn to use AI tools responsibly, they’re not just supporting library workflows—they’re preparing for the broader workforce,” he explains. For him, collaboration is professional responsibility in action: aligning institutional goals with the growth of the next generation.
Together, their dialogue reveals that collaboration around AI must operate at multiple scales. Trevor highlights national leadership, Ava stresses community and solidarity, and Russell shows how collaboration is enacted through hands-on mentorship and training. Across these levels, the message is consistent: libraries cannot navigate AI in isolation. Shared responsibility—whether through policy, partnership, or pedagogy—will determine whether AI strengthens or undermines the values that define the profession. This sense of shared responsibility sets the stage for thinking about the future—how optimism and caution combine as librarians at different career stages imagine what AI will mean for the profession in the years ahead.
Shaping the Future: Values, Not Just Tools
What emerges from this dialogue is not agreement on every point but rather a shared conviction that AI in libraries must be guided by values rather than novelty. Trevor stresses the responsibility of national leadership to build scalable frameworks that prevent inequities from widening. Ava reminds us that innovation must be tempered by harm reduction, ensuring that efficiency never eclipses justice, care, or entry points for new professionals. Russell demonstrates how these broader commitments can be enacted through institutional choices—licensing policies, workforce training, and projects that link discovery with student learning. Seen together, these perspectives suggest that AI’s impact on libraries will not be determined by the tools themselves but by the collective decisions librarians make across career stages and contexts. If guided by equity, collaboration, and critical judgment, AI can become a means of expanding access and preparing communities for the future; if not, it risks deepening divides. The challenge, then, is not simply to adopt AI but to shape it—intentionally, ethically, and in ways that preserve the human connections at the heart of librarianship. 
Note
1. Gretchen Deutsch, “California State University Bets $17 Million on ChatGPT for All Students and Faculty,” Los Angeles Magazine, August 19, 2025.
Article Views (By Year/Month)
| 2026 |
| January: 0 |
| February: 0 |
| March: 0 |
| April: 0 |
| May: 0 |
| June: 2157 |
| July: 1898 |
| August: 463 |