David Lankes, the iSchool’s Virginia and Charles Bowden Professor of Librarianship, is a busy man these days, traveling the world to discuss his recent book, Triptych: Death, AI, and Librarianship. Ahead of 2026 dates in England, Scotland, and the Netherlands, Lankes spoke to us on his way back from Sao Paolo, Brazil to expound on the vision he’s been sharing of the vital role of librarianship in the age of artificial intelligence.
Triptych takes on the epidemic of deaths of despair in the United States and asks how libraries can be part of a solution – especially in the context of emerging AI and employment upheavals. The book builds on Lankes’s long resume as a leader in envisioning the library of the future. He recently completed a funded study comparing the uses of AI at 19 state libraries around the U.S.
“This is a topic I've been talking about for a decade,” Lankes says. “It’s a conversation I've been part of around the world, looking at not only how AI will impact the field of librarianship, but how librarianship can help shape AI to be more trusted and more useful moving forward.”
Arbiters of Trust and Credibility
The title of Triptych refers to a three-paneled artwork that tells a story. Likewise, his vision for libraries the age of AI emphasizes three interlocking roles. The first role is as intermediaries between the public and AI systems, helping develop an informed understanding of when AI can be useful (and when it’s not ready), when it’s trustworthy (and when it must be taken with a grain of salt), and how it might be harnessed for our needs.
“As our communities are dealing with the rapid shifts in trust that AI is bringing on, they need a trusted agent, a trusted place where they can go to dream, dread, and deconstruct,” Lankes says. That place, he argues, should be the library.
Ordinary people need a chance to interact with AI systems – including premium, subscription-only options – and test their ambitions against these emerging capacities. They also need to experience what’s scary about AI systems, including displays of talents that might eventually replace their own human skills in the job market. Above all, people need to experiment until they fully understand what these systems are capable of.
Through all of this, people also need knowledgeable human guidance. It’s Lankes’s hope that, as libraries serve more and more as laboratories for human use of AI, that librarians are appreciated by industry as vital to discussions of AI system designs and goals. “Librarians need to be part of that conversation to meet their mission, and AI developers and builders need librarians be part of it, to talk about trustworthiness and mitigating harm,” Lankes says.
As universities reorganize resources and academic units to encourage collaboration around the emergence of AI, librarians-in-training have a key role to play, enriching the work of computer and data scientists with feedback on user trust, harm mitigation, and how well the public’s information needs are being served. “Librarians are the humans in the loop,” Lankes says.
Custodians of Data Integrity
For librarians and archivists, being “the humans in the loop” goes beyond guiding other humans confronting the AI revolution. A second key role, Lankes says, is to ensure that the data and raw materials used by advanced computer systems are of high quality – not recycled, misused, or poorly attributed slop.
“You can't run good algorithms on crappy data,” Lankes says. “And when you talk about data quality, data integrity, you're talking about the core of what libraries and archives have asked forever, both in physical form and, increasingly, in digital form: How do we assure that this data is preserved, is clean, is accessible, is useful?”
In a world where information is passed from algorithm to algorithm at inhuman speed, often subject to new interpretation and recapitulation with each new use, it can be easy to jumble sources or to confuse what exactly certain language or numbers refer to. It has long been the job of librarians and archivists to be resident experts in any community – local, academic, government, or corporate – on questions of information provenance, quality, organization, and appropriate use. Lankes anticipates this role becoming only more crucial in the coming years.
“These systems depend increasingly not just on algorithms, but on data and collections,” Lankes says. “And so how those collections and data sets and materials are organized, located, described, is essential to success.”
The value-add of librarians could go even further, into reporting back to the architects of AI systems how algorithms are being used in the real world, whether by researchers or ordinary people. This is where the field of information science brushes up against that of user experience design – both career tracks served by the iSchool, incidentally.
“On one end, the systems have to be fed, and library science and archives help with what they’re being fed,” he says. “On the other end, these systems and algorithms are being used, and knowing about the environments in which they're being used helps to influence and improve the system.”
Community Partners
Much of Triptych is concerned with the serious consequences of technological and economic upheaval on Main Street. As ordinary people try to adjust to rapid changes, the library will be one of the key places people go to adapt, whether they are seeking new employment, professional advancement, or entrepreneurial opportunity.
“It's a place where people go to do resume development, where they apply to school, where they find a space to work, where they have not only a trusted environment and trusted resources, but trusted people, in the form of librarians and archivists, to help them through that process,” Lankes says.
The experience of those affected by technological changes – both for the better (enhanced ability to pursue learning or other fulfilling pursuits) or for the worse (loss of job, sense of purpose, or social connection) – is part of the emerging story of the era of AI. Those who study people’s use of information, including on-the-ground librarians, are at the forefront of tracking this transformation and feeding findings back into building better systems.
“If you're going to talk about technologies that are going to change the culture, you’d better understand culture,” Lankes says. “How do you examine it? How do you document it? How do you enrich it?”
Above all, Lankes urges future librarians, and the schools that train them, not to succumb to a backward-looking vision based on how libraries appeared to us when we were young, as places full of treasured objects. What libraries actually do has always exceeded merely hosting collections. In the age of AI, libraries will succeed by continuing to fulfill their core mission, which is unchanged: Libraries must steward information access for the betterment of the communities they serve.
“It's okay for a 10-year-old to see the library as a bunch of stuff,” Lankes says. “But as adults and as scholars, we need to understand that its true instrumentation is as essential community infrastructure – to help people use information in a positive way, to get smarter, to make smarter decisions.”