NISO Plus 2026 Global/Online

With the growing role of agentic AI in the discovery and delivery of content, we find ourselves in a shift from human to computer readership. In this new environment, how can we embed provenance and attribution into AI systems? Without attribution and citation in outputs, how can we track usage? The session will also explore what is needed from the AI development community and how the content exchange process can support provenance tracking.

With the implementation of AI tools to support writing and editorial processes well underway, the need for policies around transparency and disclosure of AI use has grown. In response, many institutions have developed AI policies for students, researchers, and staff, while publishers are increasingly demanding disclosure of AI usage. What goes into an ethical, transparent disclosure policy? What approaches are being used to enforce such policies? After implementation are there ways we can measure their success? Society publishers, librarians, and technology providers will share their experiences and explore where advances need to be made.

Agentic AI offers new opportunities for publishers licensing content and for researchers engaged in text or data mining across a collection or corpus. Access to that content, however, also poses challenges for publishers and platforms who want to allow access to those entitled to it, whether through a subscription or license, while keeping out unwanted agents. Agentic AI also complicates how we measure the value and impact of scholarly content. What is the best approach to managing AI access to content? And how should we take AI activity into account when measuring usage?