On August 8, 2024, the International Seminar on Artificial Intelligence and Information Integrity symposium brought together global participants for thought provoking discussions at MEXINFO, hosted by UNAM in Mexico City. The seminar sought to explore the complex interplay between AI and information integrity, and to foster dialogue on the potential benefits, challenges, and ethical considerations involved.

Artificial Intelligence and Information Integrity was a collaboration of two IFLA groups: the Advisory Committee on Freedom of Access to Information and Freedom of Expression (FAIFE), and the Artificial Intelligence Special Interest Group.  We were fortunate to be part of the collaboration of IFLA groups present at MEXINFO at UNAM, which also included a seminar for Latin America and Caribbean Libraries; Unshelving Voices: Empowering Libraries against Censorship; and a symposium on Indigenous Issues in Libraries. Participants came together for the opening session, and were able to move between the symposia based on their interests.

Our keynote speaker discussing balancing the benefits and challenges of AI to information culture was Andrew Cox, convenor of the IFLA Artificial Intelligence SIG, from the University of Sheffield. Andrew shared some of the ways that AI is being used in libraries, from new ways to describe collections, to chatbots for inquiry response, to policy drafting, to storage and retrieval. He summarized some of the challenges of AI, from both an ethical and informational perspective, such as perpetuating biases, compromising privacy, and threatens trust in information systems. In addition, he shared a range of areas to consider for libraries addressing AI literacy for their users.

The program included a panel discussion of IFLA leaders and partners, reflecting how AI issues affect IFLA policy development and education activities. Jonathan Hernandez, speaking on issues affecting FAIFE; Jeannette Lebron, discussing considerations for the IFLA Latin America & Caribbean section, and Jonathan Band, speaking on copyright issues, expanded our understanding of how the unique challenges in each region affect where our attention needs to be in the library sector. In LAC, the need for an enabling policy environment and adequate funding are barriers that must be addressed before everyone in the library sector can fully participate in the potential of AI. The differences between countries was similarly important to recognize for the copyright theme, where libraries were encouraged to identify their specific copyright framework and how existing fair use/fair dealing or copyright exceptions impact the use of copyrighted content in datasets for artificial intelligence analysis, and other copyright considerations related to AI. For the work of FAIFE, the threats to information integrity are recognized globally, and the United Nation’s June publication of principles for information integrity offers opportunities for libraries to collaborate and demonstrate our roles in information and digital literacy both globally and nationally.

Participants were fortunate to hear from twelve presenters discussing specific issues related to artificial intelligence, both in-person and virtual, with themes related to information and media literacy training, awareness and use of AI in libraries and the publishing sector, impacts of AI on trust, and ethical issues in AI. Perspectives were shared from participants and studies globally, including Bangladesh, Brasil, Bulgaria, India, Italy, Mexico, Peru, Puerto Rico, Sweden, the UK, the United States, and Zimbabwe. A link to the full program is available here: https://difusion.iibi.unam.mx/mexinfo/seminarioIA-en.html

In addition, Maria De Brasdefer from IFLA HQ shared updates on IFLA’s work on digital inclusion, and the launch of the new IFLA Internet Manifesto. She noted that the manifesto provides an opportunity to renew and unify the global library position, and to provide a narrative about libraries in the digital space, as well as a way to foster partnerships and collaboration. Maria encouraged everyone to engage with the manifesto’s content and upcoming programs.

The range of views and experiences shared from across IFLA’s regions, enhanced by the combination of in-person and virtual participation, resulted in an enlightening day of programming and greater understanding of the current state of artificial intelligence use in libraries, as well as areas for our focus in the coming years.

 

Author: Christina de Castell, IFLA FAIFE Member, Information Integrity FAIFE Working Group.