Pre-conference I | Keeping the Robots in Line: Provisioning AI Access to Content
Monday, February 16 | 8:30 am–12:00 pm
Over the past year, there has been an explosion of web traffic to content sites driven by machine access ingesting content for AI tools. Publishers, repositories, content aggregators, and platform providers all have reasons—and sometimes business cases—to allow agentic AI to access their content. A publisher may choose to license some of its content to an LLM developer, or to provide subscribed access to users who want to use AI tools An open access repository may allow researchers to use AI to ingest its content and produce summaries or datasets. Whatever the case, allowing AI to access your content raises important security, systems stability, and interoperability issues, such as how to allow wanted AI agents in while keeping the unwanted out. In this session, speakers and participants will consider a range of ideas for managing AI access to content, with the goal of identifying possible best practices that work across the community to maintain service levels for humans while providing access to the robots.
Speakers
- Tony Alves, HighWire
- Todd Carpenter, NISO
- Jessica Miles, The Informed Frontier
Pre-conference II | Tracking Usage in the Age of AI
Monday, February 16 | 1:00–5:00 pm
Measuring the impact of scholarly content is critical for publishers, who must demonstrate its value to libraries and other stakeholders. Traditionally, online usage tracking was focused primarily on human activity, but machine usage of content can no longer be deprioritized; when assessing the impact of a particular title or collection, content providers and libraries are eager to capture all available data, including uses by AI. However, not all AI uses are alike: some represent a human end-user, while others originate from bots indiscriminately scraping sites for data. Stakeholders must consider iImportant questions about the different types of machine usage, such as search, summarization, and synthesis with AI tools, and how that usage might be counted and assessed differently. In this pre-conference, speakers and participants will explore how developing best practices for counting and reporting AI usage could be adapted and extended for a variety of scholarly platforms and applications. Attendees will learn about existing work as well as newly developing approaches and plan additional efforts to define AI usage for publishers and other content repositories.
Speakers
- Todd Carpenter, NISO
- Tasha Mellins-Cohen, COUNTER
- Michelle Urberg, LibLynx
Pre-conference III | JATS-Con
February 16, 9:00 am–5:00 pm
JATS-Con is a conference for anyone who uses, or is interested in learning about, the Journal Article Tag Suite (JATS), an XML format for marking up and exchanging journal content. JATS is an ANSI/NISO standard and is formally designated as ANSI/NISO Z39.96. The program will be announced soon.
Conference presentations are peer-reviewed and result in a final paper that is archived. Papers from previous conferences are available in the proceedings (https://www.ncbi.nlm.nih.gov/books/NBK65129/).