Is The Use of AI in Manuscript Editing Feasible? Here’s Three Tips to Steer Clear of Potential Issues

Presently, generative artificial intelligence (hereinafter referred to as AI) is quietly influencing almost every industry, including scientific research and academic writing. Some researchers are exploring AI-based editors to refine papers and enhance writing quality. Though AI-based editors can help authors rewrite sentences and catch grammatical errors¹, it doesn’t mean that complete reliance on AI for academic writing is advisable. Human review remains essential. 

When using AI to assist academic writing, authors must pay close attention to aspects such as contextual analysis, accuracy in expression, cultural sensitivity, and academic ethics; doing this can help avoid many common issues. Now, let’s take a look at 3 key points to keep in mind when using AI-based editors for academic writing! 

1. Analyzing context and grasping nuances in writing

Writing an academic paper requires in-depth knowledge of the subject (including recent advances in the field). Editing it also requires this context; without it an editor may change the meaning of what a paper is saying.
 

For example, consider this sentence: 

“While the researcher introduced the participants the sample images appeared one by one on the screen.” 

Here, the author could have missed adding some context and punctuation. After editing, this could read: 

“While the researcher introduced the participants to the sample images, the latter appeared one by one on the screen.” 

Or 

“While the researcher introduced the participants, the sample images appeared one by one on the screen.”
 

Choosing the option that accurately reflects the writer’s original intention requires native-level English proficiency and robust academic writing experience. An experienced human editor with a clear understanding of nuance and context would be able to do it. However, an AI-based editor may not be trained on such data. Normally, data for training AI is diverse and not subject-specific knowledge. Another consideration is that a well-written manuscript summarizes the findings of a study accurately without overstating or understating them. To meet these expectations, a thorough human review post AI editing is crucial to ensure your complex ideas come across clearly to readers.
 

2. Maintaining the sensitivity of content

AI learns based on big data, mostly generated by humans. However, all of its training data is not screened and therefore, may have biased²⁻³ or inappropriate expressions in content involving gender, race, culture, etc. AI-based editors may inherently mirror these biases. In order to avoid such mistakes, AI-edited work still needs to be carefully reviewed and screened, manually, to avoid misrepresentation in cultural and other aspects.
 

3. Providing disclosure and following strict academic ethics

Recently, accusations of plagiarism have been leveled against AI-powered articles published by some popular news channels⁴. The problem of plagiarism is even more serious in academia. In rigorous academic writing, caution is more essential when using AI. Novelty is crucial for any manuscript. Manuscripts with unoriginal ideas or language are quickly rejected by most journals⁵.  However, for the moment, AI-based editors may not catch all instances of plagiarism or copyright infringement in text, posing risks of misleading content or errors. 

Authors must understand that unpublished manuscripts uploaded for AI editing might become part of the data used to further train the AI model. Therefore, thoughtful consideration is crucial before using AI for manuscript revision or translation. 

When using AI in academic writing, authors should fully and accurately disclose its usage, apart from content control. Another reason why humans have to double-check AI-edited manuscripts is that AI cannot take responsibility for errors. Authors are solely responsible for errors made by AI-based editors⁶. This is also another reason why it is more convenient to work with human editors. 

AI-assisted technology has become more advanced in recent years, offering diverse applications. However, in the context of academic publishing, the usage of AI-based editing still requires careful judgment and review by the author. 

Editing is a fine art. It requires experience and understanding of a subject area and the nuances of academic writing. Professional language editing services can also help you effectively avoid the risks that AI modifications may bring. The Elsevier Language Services editing team, proficient across 100+ disciplines, enhances your manuscript, ensuring it aligns with the standards of top journals. Our excellent work can help increase your chances of publication. 

We also have some exciting news to share with you. The new and improved website of Elsevier Language Services is now online. It comes with a user-friendly interface and a more intuitive content layout. With improved ease of navigation, our website now guarantees researchers the ultimate browsing experience. So, enjoy our professional language services through a world-class online platform! 

References 

1. Chen, B. X. (2023, February 1). A.I. bots can’t report this column. but they can improve it. The New York Times. https://www.nytimes.com/2023/02/01/technology/personaltech/chatgpt-ai-bots-editing.html
2. Liu, R., Jia, C., Wei, J., Xu, G., & Vosoughi, S. (2022). Quantifying and alleviating political bias in language models. Artificial Intelligence, 304, 103654. https://doi.org/10.1016/j.artint.2021.103654
3. Singh, S. (2022). What you need to know about artificial intelligence in research and publishing. Editage Insights. https://www.editage.com/insights/what-you-need-to-know-about-artificial-intelligence-in-research-and-publishing
4. Tameez, H. (n.d.). BuzzFeed will start using AI to write quizzes and other content. Nieman Lab. https://www.niemanlab.org/2023/01/buzzfeed-will-start-using-ai-to-write-quizzes-and-other-content/
5. Ali, J. (2010b). Manuscript rejection: Causes and remedies. Journal of Young Pharmacists, 2(1), 3–6. https://doi.org/10.4103/0975-1483.62205
6. Thorp, H.H. (2023). ChatGPT is fun, but not an author. Science, 379(6630). https://doi.org/10.1126/science.adg7879.
7. Elsevier. (n.d.-c). The use of generative AI and AI-assisted technologies in writing for Elsevier. https://www.elsevier.com/about/policies/publishing-ethics-books/the-use-of-ai-and-ai-assisted-technologies-in-writing-for-elsevier