Artificial intelligence (AI), particularly Generative AI (like ChatGPT) and Predictive AI (machine learning algorithms), is rapidly changing the landscape of scholarly research and publishing. As scholars, it is essential to understand both the opportunities and the ethical challenges these tools present.
The core rule for using AI in scholarly work is to approach it with a critical lens and know the policies of your chosen publisher.
Check Journal Policies: Always consult the journal or publisher's specific policy regarding the use of AI in the creation of the submitted manuscript. Policies can be vague, but adherence is mandatory.
The Author's Role: Even if AI assists with writing, the faculty member remains the sole author and responsible party for the entire work, including the accuracy and originality of the content. If you use AI for any significant portion of writing (even polishing), disclosure may be required, though authors often fail to do so despite policies. Authors who falsely represent their use of AI tools risk having their articles retracted if their deception is discovered.
AI Cannot Be an Author: Most major publishers and editorial organizations prohibit listing a generative AI tool (like ChatGPT) as an author, as it cannot take responsibility for the content. If you want your scholarly writings to be protected by copyright, it is essential that you do the writing, not AI.
AI in Peer Review: While some journals are experimenting with AI to assist in peer review, many editorial organizations have guidelines about the ethical use of these tools and may require reviewers to be transparent about any such use. Some (like NIH) prohibit their use in reviewing. AI for screening and peer review is often problematic and ineffective despite common use by publishers.
Detection: Be aware that "tortured phrases"—unnatural language created by generative AI—can be apparent to editors and reviewers, similar to signs of plagiarism.
Bias and Inaccuracy: AI systems often reinforce and replicate biases present in their training data. Additionally, AI may hallucinate citations or generate false information, especially in highly specialized fields like math or law. Always cross-check generative AI results.
Copyright and Intellectual Property: There are ongoing concerns among authors and publishers about the unauthorized, uncompensated training of copyrighted content for Large Language Models (LLMs). If you’re concerned by the possibility that your scholarly work could be sold to train AI, be sure to read and understand the terms of your future publication contracts, both for articles and for books.
While AI tools can do many impressive things, they're also surprisingly unreliable in critical areas. For example, AI tools are notorious for generating false information, and that these “hallucinations” extend to citations as well. Many AI tools spit out seemingly reasonable citations to journal articles and other scholarly works, complete with real author names and plausible titles, that are utterly fictitious. Some AI tools are specifically designed for research and can reliably generate citations that are both real and relevant, but using these tools often requires a paid account.
Similarly, many AI tools can summarize text or list key findings from publications. These summaries and distillations are typically convincing at first glance, but they often reveal errors, biases, or distortions upon closer inspection. If you employ AI tools when searching for scholarly works or otherwise working on your literature review, be sure to also investigate the limitations of those tools, examine their output closely, and double-check any facts or citations they provide. Failing to do so could result in embarrassing and potentially career-damaging flaws in your own work.
When used judiciously and critically, AI tools can be a significant time-saver in the research process.
| Level | Prompt |
| (1 ) Basic editing, such as checking spelling and grammar, or suggesting synonyms. | Check the spelling and grammar in this paragraph, and suggest synonyms for any repetitive words. |
| (2) Structural editing, such as paraphrasing, translating, or improving the structure of the text, or its flow or coherence. | Paraphrase this lengthy sentence to improve its clarity and flow, and translate it to French. |
| (3) Creating derivative content, such as summarizing, creating titles and abstracts, rewriting or generating analogies. | Summarize this document and create a short, catchy title for a journal submission. |
| (4) Creating new content, such as completing, continuing or expanding text, or brainstorming ideas. | Continue the text to explain the key question being addressed. Show why it is important, drawing parallels or analogies where you see fit. |
| (5) Evaluation or feedback, such as assessing the quality of the writing or finding weaknesses in it. | Review this introduction and highlight any logical gaps or areas that need further development. |
Tools for polishing: Grammarly, Paperpal, Writefull, and Curie can provide feedback on small sections of writing.
*Lin, Z. (2024). Techniques for supercharging academic writing with generative AI. Nature Biomedical Engineering, 9(4), 426–431. https://doi.org/10.1038/s41551-024-01185-8
Adapted from:
AI for research and scholarly publishing by OpenLab at City Tech
AI & Scholarly Publishing by Jill Cirasella, Scholarly Communication Librarian and University Liaison, CUNY/The Graduate Center
