What's Happening?
COUNTER (Counting Online Usage of Networked Electronic Resources) has updated its Best Practice on Generative and Agentic AI Usage Metrics, effective October 23, 2026, to differentiate between content retrieved by AI systems and content actually cited
or used in an answer. This update addresses the evolving landscape of scholarly content consumption in the age of AI, where traditional download metrics no longer fully capture how AI assistants interact with academic literature. The new guidance specifies that an AI system's retrieval of excerpts counts towards retrieval metrics, even if those excerpts are not ultimately selected for an answer. Conversely, a cited or presented item may contribute to an AI Investigation or AI Request if the full text is accessible. COUNTER's framework, which maps events from the SPUR Coalition's telemetry specification, aims to provide a more granular and accurate understanding of AI-mediated scholarly content use, recognizing that a thousand excerpt retrievals do not equate to a thousand human readers or articles.
Why It's Important?
This update is crucial for publishers, libraries, and researchers in the U.S. and globally, as it provides a standardized method for measuring AI's engagement with scholarly content. For publishers, it offers a clearer picture of the value and reach of their content beyond traditional human-driven downloads, which are expected to decline as AI adoption grows. Libraries can use these refined metrics to make more informed decisions about subscriptions and resource allocation, understanding how their licensed content is being utilized by AI systems to support research. The distinction between retrieval and citation helps to prevent inflated usage figures and promotes transparency in reporting. This move by COUNTER, in coordination with initiatives like SPUR and NISO, is vital for establishing a robust and auditable framework for AI-mediated scholarly research, ensuring that the impact and contribution of academic work are accurately recognized in the AI era.
What's Next?
The implementation of COUNTER's updated guidance will lead to changes in how AI-mediated scholarly content usage is reported and analyzed. Publishers will need to adapt their reporting systems to align with the new metrics, providing more detailed breakdowns of AI interactions. Libraries will begin to incorporate these new data points into their collection development and budget allocation processes. Further collaboration among organizations like COUNTER, SPUR, NISO, and TRACE (Trusted Retrieval and Attribution for Content Ecosystems) is anticipated to refine these metrics and develop additional standards for provenance and attribution in AI-generated content. The industry will likely continue to explore how to measure the 'contribution' and 'engagement' aspects of AI use, beyond mere retrieval, to fully understand the value generated by scholarly content in AI workflows. This ongoing evolution aims to ensure fair commercial relationships between publishers and AI providers.
Beyond the Headlines
Beyond the immediate impact on usage metrics, this development touches upon fundamental questions regarding the value of scholarly content in an AI-driven world. It highlights the challenge of attributing intellectual contribution when AI systems synthesize information from numerous sources. The distinction between retrieval and citation underscores the need for AI models to provide transparent provenance and attribution for the information they present. This also raises ethical considerations about the 'black box' nature of some AI systems and the potential for unacknowledged use of copyrighted material. As AI becomes more integrated into research, the debate over how to fairly compensate content creators and ensure the integrity of scholarly communication will intensify. The efforts to standardize AI usage metrics are a critical step towards establishing a sustainable ecosystem where both human and artificial intelligence can thrive responsibly.













