Navigating AI in Academic Writing: Guidelines and Best Practices

Artificial intelligence (AI) refers to intelligent machines and algorithms that can reason and adapt based on sets of rules and environments, mimicking human intelligence. This field is evolving rapidly, and the education sector is abuzz with discussions on AI use for writing. This matters not just for academics but for anyone relying on trustworthy information, from journalists and policymakers to educators and the public. Ensuring transparency in how AI is used protects the credibility of all published knowledge.

In education and research, AI can generate text, improve writing style, and even analyze data. It saves time and resources by allowing quick summarizing of work, language editing, and reference checking. It also holds potential for enhancing scholarly work and inspiring new ideas. Equally, AI is capable of generating entire pieces of work, sometimes making it difficult to distinguish original work written by an individual from work generated by AI.

This is a serious concern in the academic world, affecting universities, researchers, lecturers, and students. Some uses of AI are seen as acceptable, while others are not. As an editor and editorial board member of several journals, and in my capacity as a researcher and professor of psychology, I have grappled with what counts as acceptable use of AI in academic writing. I looked to various published guidelines for insights.

Assisted versus Generated Content

In understanding AI use in academic writing, it’s important to distinguish between AI-assisted content and AI-generated content. AI-assisted content refers to work that is predominantly written by an individual but has been improved with the aid of AI tools. For example, an author might use AI to assist with grammar checks, enhance sentence clarity, or provide style suggestions. The author remains in control, and the AI merely acts as a tool to polish the final product. This kind of assistance is generally accepted by most publishers without the need for formal disclosure, as long as the work remains original and the integrity of the research is upheld.

AI-generated content, on the other hand, is produced by the AI itself. This could mean that the AI tool generates significant portions of text or even entire sections based on detailed instructions provided by the author. This raises ethical concerns, especially regarding originality, accuracy, and authorship. Generative AI draws its content from various sources such as web scraping, public datasets, code repositories, and user-generated content. AI “hallucinations” are common, and generative AI might be plagiarizing someone else’s work or infringing on copyright without the author’s knowledge.

What’s Allowed and What’s Not

Based on my readings of the guidelines, I offer some practical tips for using AI in academic writing:

• AI tools can be used for routine tasks like improving grammar, revising sentence structure, or assisting with literature searches. These applications do not require specific acknowledgement.
• Across the guidelines reviewed, AI-generated content is not allowed unless there are clear reasons why this was necessary for the research and the content is clearly marked and referenced as such. Thus, depending on how AI is used, it must be referenced in the manuscript. This could be in the literature review, methods, or results section.
• Authors must disclose when AI-generated content is used by citing this appropriately. The name of the generative tool used, the date accessed, and the prompt used should be cited. This level of transparency is necessary to uphold the credibility of academic work.
• Other aspects linked to AI assistance, like correcting code, generating tables or figures, reducing word count, or checking on analyses, cannot be referenced directly in the body of the manuscript. In line with current best practice recommendations, this should be indicated at the end of the manuscript.
• Authors are responsible for checking the accuracy of any AI content, whether AI-assisted or AI-generated, ensuring it’s free from bias, plagiarism, and potential copyright infringements.

The Final Word (for Now)

AI tools can undoubtedly enhance the academic writing process, but their use must be approached with transparency, caution, and respect for ethical standards. Authors must remain vigilant in maintaining academic integrity, particularly when AI is involved. Authors should verify the accuracy and appropriateness of AI-generated content, ensuring it doesn’t compromise the originality or validity of their work. There have been excellent suggestions regarding when the declaration of AI should be mandatory, optional, or unnecessary. If unsure, the best advice would be to include the use of any form of AI (assisted or generated) in the acknowledgment. AI tools are here to stay. Let’s deal with it constructively and collaboratively.