R0029/2026-03-27/Q003/SRC06/E01¶
Big 5 publisher AI policy comparison: prohibition consensus, disclosure variation
URL: https://www.thesify.ai/blog/ai-policies-academic-publishing-2025
Extract¶
REPORTED: Comparative analysis of Big 5 publisher AI policies:
| Publisher | AI Authorship | Disclosure Location | Grammar Exempt | AI Images |
|---|---|---|---|---|
| Elsevier | Prohibited | Dedicated AI declaration | Yes | Prohibited |
| Springer Nature | Prohibited | Methods section | Yes (copy editing) | Prohibited (exceptions) |
| Wiley | Prohibited | Methods or Acknowledgments | Yes | Prohibited |
| Taylor & Francis | Prohibited | Methods or Acknowledgments | Not explicit | Prohibited |
| SAGE | Prohibited | Formal citation + prompts | Assistive AI exempt | Implicit restriction |
All five agree on: prohibition of AI authorship, human accountability, and some form of disclosure. They differ on: where to disclose, what triggers disclosure, how to handle AI images, and peer review AI use.
Relevance to Hypotheses¶
| Hypothesis | Relationship | Strength |
|---|---|---|
| H1 | Supports | All five prohibit and require disclosure |
| H2 | Contradicts | All five have formal policies |
| H3 | Supports | Five different disclosure approaches from five publishers |
Context¶
The variation across five publishers that compete in the same market suggests the field has achieved principle-level consensus (no AI authors) but not implementation-level standardization.