The AI Contribution Matrix: A Standardized Framework for Disclosing AI Use in Scientific Papers
Year: 2026
Type: preprint
Abstract: Generative Artificial Intelligence (AI) is changing how research is conducted, with AI being increasingly used for writing papers. However, only a small minority of papers disclose that AI was used, let alone how. This leads to a significant lack of transparency, which could result in a loss of research integrity. Taking into account the perspectives of authors, readers, and reviewers, we have developed a guideline for disclosing the use of AI in scientific papers. At its core, the guideline introduces the AI Contribution Matrix -a standardized, research-discipline-independent framework for reporting AI usage in a structured manner. The AI Contribution Matrix consists of six dimensions: Use Case, AI System, Date, Contribution, Review, and Comment. Furthermore, guidance is provided on mandatory and voluntary disclosures, when multiple authors are involved in a paper, where the AI Contribution Matrix should be placed, and what to do if no AI was used in a paper. Additionally, we present options for how outlets can individually adapt the guideline on permitted and undesired AI usage.
Source: HAL
Authors: Witte, Jonas, Weber, Ingo
Institutions: Fraunhofer-Gesellschaft, Technical University of Munich
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7169479764