AI evidence extraction

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

Emily Genatowski · Preprints.org · 2026

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

9/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.20944/preprints202602.1215.v1

Methodology & findings

Study design

Mixed-method approach combining: (1) Two interdisciplinary workshops with approximately 10 attendees across academic fields discussing six citation categories; (2) Anonymous poll of graduate students at University of Vienna's AI in Academia Workshop; (3) Presentations and discussions at International Love Data Week (University of Graz), Emerging Digital Methodologies Conference (Oxford University), and guest lectures in a Digital Humanities reading course at University of Vienna; (4) Synthesis of notes from discussions into proposed citation conventions..

Main result

The paper presents seven citation templates and a flexible framework for AI-supported scholarship. The authors found that "The variations occur in how the prompts are engineered e.g. single prompt, multimodal prompt or multi turn refinement, the application of the output in the academic work e.g. citation of output text or as a tool or method, and how much agency the AI support has in the creation of the scholarly work e.g. authorship or co-collaborator." The workshop discussions emphasized prompt transparency, archiving requirements, model version identification, and established a policy that "AI cannot be listed as a co-author."

Research paradigm

Interpretivist/qualitative - developing normative standards through interdisciplinary discussion and stakeholder engagement

Author conclusions

The authors conclude: "AI is no longer peripheral to research, but citation standards still continue to lag behind. Without clear norms, AI remains invisible. With them, AI use has the potential to become transparent and accountable. The CLARIAH-AT project demonstrates rigorous, flexible, adaptable and usable standards are possible. Seven templates and an archival suggestion offer a prospective roadmap for scholars and institutions as well as continue the discussion surrounding AI acceptance and usability in the scholarly context." They further state: "Embracing these temporary formatting suggestions will give historians, librarians, and students the confidence to work openly with AI."

Risk of bias

Selection bias: approximately 10 self-selected workshop participants across undefined institutions may not represent broader scholarly community; Institutional bias: project originated from University of Vienna; primary participants appear concentrated in European academic institutions; Author bias: project co-leads (Genatowski and Wallnig) directly facilitated workshops and synthesized notes, introducing potential confirmation bias in interpretation; Disciplinary bias: workshop attendees described as 'across academic fields' but specific disciplines not enumerated; may skew toward digital humanities/STEM fields; Selection bias in workshop participants (self-selected academics interested in citation standards); Non-representative sample size (approximately 10 workshop attendees); Limited geographic scope (primarily Austria-based institutions); Disciplinary bias (overrepresentation possible from humanities/digital humanities fields given CLARIAH-AT context); No information on graduate student poll sample characteristics or representativeness

Open questions raised

The paper identifies several gaps: (1) lack of citation conventions for AI tools in research workflows; (2) inadequacy of traditional citation formats for dynamic, personalized AI outputs; (3) misalignment between international citation standards (APA, MLA, Chicago) treating AI differently; (4) gap between legal frameworks defining AI output ownership and institutional policies on authorship attribution; (5) inconsistencies between institutional guidance and departmental practices on AI citation; (6) need for persistent identifiers (PIDs) for AI models and outputs to ensure traceability.
Development of international unified standards and interoperability through DOI and ORCID registries
Reconciliation of legal frameworks regarding AI output ownership with institutional attribution policies
Integration of persistent identifiers (PIDs) for AI models and outputs into citation metadata and FAIR principles
Development of dynamic institutional policies that bridge departmental expectations and disciplinary norms
Ongoing evolution of citation standards as technology continues to develop

Extracted from: pdfAgreement 67%

Explore related topics

Citation Integrity Scholarly Infrastructure Position Paper