Generative AI and peer review: Summary

Between December 2023 and the early months of 2024, I denounced a language journal in Italy, alleging that the two peer review reports for my submitted article had been produced using generative AI (ChatGPT). I pursued this matter publicly in the equal aims of establishing accountability and of raising public awareness about the new danger LLMs pose to scholarly integrity in the AI era.

‍

Those blog posts have been archived are no longer being updated. They are accessible here:

• “Unethical academics, AI, and peer review,” 20 December 2023 [link]
• “Updates on the AI peer review incident,” 31 January 2024, 7 February 2024, 6 April 2024 [link]
• “AI and peer review: Official journal response,” 26 February 2024 [link]‍

‍

Read about this incident in my case report published in the open-access, open-peer-review journal Research Integrity and Peer Review:

• Nicholas Lo Vecchio. 2025. “Personal experience with AI-generated peer reviews: A case study.” Research Integrity and Peer Review 10:4. https://doi.org/10.1186/s41073-025-00161-3 [link]

‍

Slides from my talk at the Sorbonne in November 2025:

• Nicholas Lo Vecchio. 2025. “L’intrusion de l’IA dans le monde académique. Réflexions suite à une expertise automatisée par chatbot.” Conférence à Sorbonne Université, 25 nov. 2025. [link]

‍

Write-up in Mediapart about the Toulouse conference in June 2026, where this subject among many others was addressed:

• Livia Garrigue, "Objection de conscience, pragmatisme ou déni: les universitaires face à la déferlante de l’IA,” Mediapart, 5 July 2026 [lien]
• Conference "Que faire face à l’IA générative dans l’enseignement supérieur et la recherche?" organized by the Atelier d’écologie politique (Atécopol), Toulouse, 18 June 2026 [lien]

‍

Documentation‍

The following pieces of supporting evidence are available here for download:

• Original copy of the peer review reports
• Author-annotated copy of the peer review reports
• ChatGPT simulations (synthesis document used as corpus for concordances)
• ChatGPT simulations raw output
• Concordances of ChatGPT simulations and peer review reports via Sketch Engine‍
• mediAzioni journal official response of 22 February 2024

‍

Author position on the LLM output in the above documentation: I oppose engaging with machine-generated content unsituated with respect to any individual human taking responsibility for it. I reject all of the ChatGPT recommendations in the documents linked as they are merely text sequences statistically reproducing normative text patterns in the corpus. I provide this documentation merely as concrete comparisons to contextualize the peer review reports I received.

‍

Nicholas Lo Vecchio‍

9 January 2025