Reproduce “AI-Generated Figures in Academic Publishing: Policies, Tools, and Practical Guidelines” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Davie Chen
Davie Chen from the University of Arts in Poznań systematically surveyed academic publishers' policies on AI-generated figures, analyzing concerns like reproducibility and visual misinformation, and subsequently developed practical guidelines for transparent and compliant use of these tools. The research identified universal disclosure requirements and AI authorship prohibitions, alongside a spectrum of varying restrictions across major journals.
Key insights
- Publisher policies universally mandate disclosure of AI tool usage and prohibit AI from being listed as an author, yet exhibit significant variation in actual restrictions on AI-generated figures.
- The primary concerns revolve around the potential for visual misinformation, especially in 'data figures,' and challenges in reproducibility due to the stochastic nature and opacity of AI models.
- Domain-specific AI tools for scientific illustration offer advantages over general-purpose generators, providing better text rendering, consistent visual styles, and structured metadata for academic disclosure.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.