Disclosure in the era of generative artificial intelligence

Key Points

• To address the inconsistencies in disclosure expectations surrounding generative artificial intelligence in academic writing.
• Analyzed current disclosure practices in several journals regarding AI usage in manuscripts.
• Proposed a framework that separates disclosure requirements based on the risk level associated with AI's influence on content.
• Identified that many journals utilize binary disclosure statements without nuanced differentiation.
• Recommended that disclosures reflect the degree of AI assistance to promote transparency in academic publications.

Abstract