NISO Plus Leans Heavily into Our AI Standards Needs

Letter from the Executive Editor, March 2026

Several times over the span of my career, there have been periods of rapid change across the community punctuated by a flurry of activity and serious considerations of strategic priorities in light of the newly altered environment.  It very much feels like we are amid another one of those periods. Fortunately, the February NISO Plus conference in Baltimore was for three days an engaged epicenter of these technical conversations around AI systems for the publishing and library communities.

In Baltimore there were dozens of conversations about how our community should address the myriad issues related to artificial intelligence systems. From reporting on third-party agentic usage, to tracking ingest and citation using the relatively new International Standard Content Code, to supporting interoperability through the Model Context Protocol, the topic of AI systems and how to support interactions with them was—by far—the most discussed during the sessions. The brilliant keynote speaker Brandie Nonnecke kicked off the meeting wonderfully as she delved into how these systems work, how they should be responsibly deployed, and how critical it is for values to be embedded in an organization’s AI strategies. Among the suggestions she advanced was the development of sector-specific training opportunities on AI governance and implementation strategies.  

The broader impacts of technology and social values were the focus of this year’s Miles Conrad Lecture given by Dr. Alondra Nelson. Dr. Nelson seamlessly connected NISO’s and NFAIS’s foundational work to improve information distribution through collaboration and standards, her own research, and her work at the White House’s Office of Science and Technology Policy (OSTP) in the Biden administration. Both Nonnecke’s and Nelson’s presentations are freely available in the NISO Video Library, and I strongly encourage you to listen to their insights and advice for our community.

Because NISO Plus is all about conversations that lead to practical outcomes, it is worth mentioning a few outcomes that are getting underway.  COUNTER Metrics just closed a public consultation on AI usage metrics. This was discussed by COUNTER Executive Director  Tasha Mellins-Cohen during the NISO Plus pre-meeting session on Tracking AI Usage. The consultation proposed extensions to the Code of Practice by adding a reserved Access_Method of Agent and adding a new a Metric_Type for measuring total and unique uses of content by AI tools. This is however, only a small slice of a considerably more complex question about how AI tools are engaging with publisher’s systems and content. A recent post on The Scholarly Kitchen by Michelle Urberg and Chris Bendall highlighted the complexities of measuring AI usage and noted that AI agents are accessing content more frequently and that search is increasingly happening in a zero-click environment. Grow Kudos is also currently undertaking an investigation of this situation for academic publishers. To address some of these larger questions, NISO and COUNTER, with the support of Cambridge University Press, will host a targeted, invitation-only event at the Cambridge University campus in May to further discuss these issues and plan for future work regarding AI usage. Look forward to a report of that workshop and its outcomes later this spring.

Training was another theme that came up in the conference: people need to expand their knowledge and skills relating to these systems. Building on the theme of AI, it is not too late to sign up for the upcoming AI training series that NISO will host starting later this month.  Led by Jessica Miles, founder of The Informed Frontier consultancy, the eight-week course will help support your team in assessing AI tools, building essential AI literacy grounded in integrity and ethics, and exploring how AI is reshaping scholarship.                  

There were many more suggestions of potential consensus efforts that were discussed during the meeting. Just a few of the early suggestions were mentioned and discussed in a closing session led by NISO’s Board Chair, Jonathan Clark from the DOI Foundation:

• Advance work on interoperability and management of AI agent systems and how they access content
• Create a taxonomy of bots and/or bot behavior
• Develop a community best practice for death date and verified affiliated researcher/institution associated with the deceased
• Explore making rights and licensing metadata both machine-actionable and enforceable, rather than just advisory
• Engage in community development of scholarly MCP applications
• Explore standards related to transparency and disclosure of AI use
• Organize a community of interest for AI tool developers in the scholarly community

The NISO staff, leadership committees, and broader community members will be going over the extensive collaborative notes documents (linked from the agenda pages) to identify and prioritize potential work that NISO or other community groups could advance. If anything catches your eye and you feel compelled to push one of these ideas forward, please reach out to me. As I’ve said every year at NISO Plus, “The real work of the conference begins the day the meeting ends.”  Please join us in bringing that work to life!

Sincerely,

Todd A. Carpenter