Do Ophthalmology Journals Have AI Policies for Manuscript Writing?

Purpose: To assess the prevalence of artificial intelligence (AI) usage policies in manuscript writing in PubMed-indexed ophthalmology journals and examine the relationship between the adoption of these policies and journal characteristics.

Design: Cross-sectional study.

Subjects: PubMed-indexed ophthalmology journals.

Main Outcome Measures: Prevalence of policies in journal guidelines regarding the use of AI in manuscript writing.

Methods: We reviewed the guidelines of 84 ophthalmology journals indexed in PubMed to determine the presence of AI-use policies for manuscript generation. We further compared journal metrics, such as CiteScore, Journal Impact Factor (JIF), Journal Citation Indicator (JCI), Source Normalized Impact per Paper (SNIP), and SCImago Journal Rank (SJR), between journals with and without AI policies. Additionally, we analyzed the association between AI policy adoption and journal characteristics, such as MEDLINE indexing and society affiliation.

Results: Among the 84 journals, 53 (63.1%) had AI policies for manuscript generation, with no significant changes observed during the study period. Journals indexed in MEDLINE were significantly more likely to have AI policies (68.8%) than non-MEDLINE-indexed journals, where no AI policies were found (0%) (P = .0008). There was no significant difference in AI policy adoption between society-affiliated (62.7%) and unaffiliated journals (64.7%) (P = .84). Journals with AI policies had significantly higher metrics, including CiteScore, SNIP, SJR, JIF, and JCI (P < .05).

Conclusions: While many ophthalmology journals have adopted AI policies, the lack of guidelines in over one-third of journals highlights a critical need for consistent and comprehensive AI policies, particularly as the AI landscape rapidly advances.