Citing the Uncitable: Developing Standards for AI and New Media in Scholarly Work

Artificial intelligence and generative models are rapidly reshaping academic research methodologies, yet cohesive citation standards remain underdeveloped. Without reliable, transparent, and standardized practices accepted by the scholarly community, AI-supported research risks being either underreported or deemed methodologically untrustworthy. The citation of AI use raises foundational challenges, particularly because AI outputs are neither fully replicable nor verifiable—two core tenets of traditional citation practice. Outputs from large language models are non-deterministic, ephemeral, and context-sensitive, necessitating renewed engagement with debates on model variability, training data, hosting environments, prompting practices, authorship, licensing, ethics, and archival responsibility. This paper presents findings from a CLARIAH-AT–funded initiative to develop citation standards for six categories of new media, with a focused examination of AI outputs. Drawing on interdisciplinary workshops, teaching practice, and scholarly discussion, the project proposes flexible citation frameworks and use-based variations that align emerging technologies with established principles of historical source criticism. These efforts demonstrate that while the citation of AI and new media remains complex and contested, transparent methodological standards can serve as an ethical and practical roadmap for scholarship in the age of AI.