Exploring AI futures in libraries with the Futurescape Workshop community

Author: Robert Hilliker

Director of Library Relations

Following my reflections from ALA 2025, where the evolving role of librarians in research management took centre stage, I recently had the opportunity to explore another frontier shaping our profession: artificial intelligence. At the Futurescape Libraries AI Workshop hosted by Carnegie Mellon University, librarians from across North America gathered to critically examine how AI might reshape discovery, access and trust in scholarly information. This column builds on the themes of partnership and proactive engagement I discussed previously, turning our attention to how librarians can help steer AI development toward values-driven outcomes.

In September, I attended the Futurescape Libraries AI Workshop under the auspices of the Association of Research Libraries (ARL) and the Coalition for Networked Information (CNI) in Pittsburgh. The workshop was grounded in the newly released toolkit designed to walk librarians through an exploration of the AI scenarios released by ARL and CNI in June 2024. There’s a lot of speculation and debate about AI right now. The workshop offered a welcome chance to dig deeper than the headlines and reconnect with fellow librarians. It gave us space to think critically about the range of possible futures ahead. We also explored what we, as librarians, can do to support each other, our patrons, and our institutions in preparing for those futures, both the promising and the challenging ones.

If you’re not familiar with the scenarios, I encourage you to check them out for yourselves, but I will summarize them briefly here (paraphrasing from the report):

• Scenario 1: Democratized and Socially Integrated AI, where the key players in government, industry, and education have come together to develop a robust and equitable framework for responsibly integrating the benefits of AI into all aspects of human society;
• Scenario 2: Consumer-Oriented AI Focused on Education and Entertainment, where large technology companies lead the field of AI development and focus on profitable direct-to-consumer applications;
• Scenario 3: Laissez-Faire AI, where widespread adoption of AI and an absence of functional regulation result in an erosion of trustworthy information, amplification of bias, and deliberate manipulation by bad actors;
• Scenario 4: Autonomous AI, where the scarcity and rising cost of human expertise leads key players to focus on developing AI to operate independently from human input, creating a world in which the benefits of AI are realized by ceding human agency.

Exploring the cone of possibilities for AI in libraries

As we were reminded at the beginning of the workshop: the purpose of these scenarios is not to predict which one will become our future reality, but to help us explore the “cone of possibilities” and identify the levers in our current reality that we can push on to try to steer away from the worst outcomes and towards the better ones. 

“For me, this open-ended and values-oriented approach to thinking about AI was a welcome tonic to the frequent hype and presentism that can dominate social media conversations about the wide range of technologies that are shoehorned under the capacious (and sometimes misleading) AI label.” 

And I found that my fellow attendees, who came from a wide range of libraries around North America, had a lot of insight to share about the challenges libraries are already facing as a result of new AI technologies, but also some of the excellent opportunities it offers us.

How AI is reshaping discovery in academic libraries

Interestingly, a lot of the challenges and opportunities AI brings to libraries seem to center around discovery. On the plus side, AI tools are getting really good at describing large digital archive collections. That means researchers, students and the public can access materials that were previously hard to find. Projects like the “Exploring Computational Description” at the Library of Congress and Yale’s “Digital Collections AI” show how fast these human-in-the-loop tools are improving. On the other, AI is being used to generate mediocre and, frequently, misleading content in volumes that threaten to swamp our human capacity for evaluation and curation. 

This use of AI poses a special problem for scholarly publishers as well, and Springer Nature is actively investing in (and sharing with the publishing community) AI tools to help identify and counter the rise of AI-driven misinformation in the scholarly publishing ecosystem.”

In fact, one of my big takeaways from this workshop was that there is already significant agreement between scholarly publishers, libraries and research institutions on the need to bring our shared, human-centered values to guide us towards a responsible use of AI. Transparency, accountability, fairness and dignity, coupled with strong data governance and a concerted effort to prevent harm to society and the environment, these are at the core of Springer Nature’s AI Principles. They align closely to the values expressed in ACRL’s AI Competencies for Academic Library Workers, particularly those expressed in the “Ethical Considerations” section of that document. The ACRL’s AI Competencies highlight the need to understand how different AI tools work and what they’re used for. They also encourage librarians to help patrons figure out when and how to use these tools to meet their information needs.

Strengthening scholarly trust through better AI tools

Clearly generative AI has added to the size of this challenge, and I was pleased to see attendees of the workshop lean into thinking about ways to address it. Ultimately, if researchers, librarians and academic publishers want to strengthen trust with the broader public, we need to rethink how we use AI. That means finding ways to give users better access to reliable information. Our discovery tools should meet people where they are.

“Everyone of us attending the workshop was under no illusions about the scope of the challenges academic libraries face right now and how to increase trust in scholarly sources of information is one of the greatest.”

At the same time, discovery tools need to do more than just deliver information. They should encourage users to think critically. That means prompting them to look at key sections of source texts and compare different perspectives. In other words, we need to support critical information literacy, so users can judge the quality and accuracy of what they find. This means being innovative in the interfaces we design while ensuring that the underlying architecture is solid, and that the technical and descriptive metadata we generate is robust and standards-compliant to support interoperability.

It’s a big challenge, but the good news is that several initiatives are already underway. Groups like STM, COUNTER, and NISO are working on the technical details, while keeping shared values front and center. And while many librarians harbor some skepticism about AI, it's important to note that being able to ask tough questions about how we can harness AI well is actually a strength. You’re already practicing one of the ACRL’s AI competencies: critical engagement!  So I encourage you to learn more about these efforts. Your constructive engagement will help make the solutions we build even stronger.

If you're interested in how the broader scholarly ecosystem is responding to AI, I encourage you to explore Perspectives on AI in scholarly communications, a report that brings together voices from several library leaders and Springer Nature experts reflecting on the opportunities and challenges ahead.

Related content

• Building better research support through community insight
• How librarians help researchers maximise their reach
• The inside out librarian – How libraries are becoming essential research partners
• Life in the Library – Building a career in academic librarianship
• Future-proofing academic librarians with essential skills and leadership
• Evolving roles of academic librarians adapting to change and new technologies
• Practical tips for librarians to support early career researchers’ success

Don't miss the latest news & blogs, subscribe to The Link Alerts!