What is it about?

This article argues that concerns about AI-assisted writing in research may be aimed at the wrong target. The author suggests that scholarly integrity depends on researchers taking responsibility for their ideas and findings, not on whether they use AI tools to help draft text. Instead of threatening academic communication, generative AI exposes weaknesses in how scholarly quality is assessed. The article introduces the idea of “efficiency shame” to describe the growing discomfort researchers may feel when using effective AI tools, and calls for a rethinking of peer review and research evaluation rather than restrictions on responsible AI use.

Featured Image

Why is it important?

Generative AI is rapidly changing how research is written, reviewed, and published. As journals and universities develop policies for AI use, it is important to distinguish between the use of AI as a writing tool and the responsibility researchers have for the claims they make. The article argues that the real challenge lies not in AI-assisted writing itself, but in how academic institutions evaluate research quality. By highlighting the concept of “efficiency shame”, it raises important questions about whether current publishing practices are adapting appropriately to new technologies and how scholarly communication can remain rigorous, transparent, and accountable in the age of AI.

Perspectives

This article grew out of my frustration with emerging AI policies in academic journals. While the idea of “efficiency shame” applies far beyond scholarly publishing, I encounter it daily as both an author and an editor. Drawing on my background in Speech Act Theory, I began to question whether we were focusing on the right problem. Generative AI is here to stay. The cat is out of the bag, and we are not going back to quill and parchment. Why would we?

Pär Ågerfalk
Uppsala Universitet

Read the Original

This page is a summary of: From Integrity to Efficiency Shame: Rethinking Generative AI in Scholarly Publishing, interactions, August 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3833436.
You can read the full text:

Read

Contributors

The following have contributed to this page