We understand that many researchers are now using AI-powered assistants to help streamline their writing and brainstorming process. At James Nicholas Publishers, we see the value in these tools which can be excellent for optimising edits and finding fresh perspectives. That said, efficiency shouldn’t come at the cost of integrity.

James Nicholas Publishers remains committed to the unique value of human authorship. Since Large Language Models (LLMs) can’t be held accountable for their output, they cannot be listed as authors on any submission. Ultimately, the responsibility for the accuracy of the work rests with you, the human author. These guidelines were created to help you use generative AI responsibly while maintaining the high standards of our journals.

Assistive AI tools

Assistive AI tools make corrections, suggestions, and improvements to content you have already written yourself. For many years we have been using tools that have offered to flag spelling or grammatical errors. Recently, these tools have introduced features to make suggestions for the next word or phrase or to suggest better or more concise phrasing to improve clarity. Content that you have written on your own, but refined with the help of this kind of Assistive AI tool is considered “AI-assisted”.

Generative AI tools

This term refers to one of the many new tools such as ChatGPT or Google Gemini which can produce content in several forms including text and images. Even if you’ve made significant changes to the content afterwards, if an AI tool was the primary creator of the content, the content is considered “AI-generated”.

Disclosure

AI-assisted writing is becoming more prevalent through everyday word processors. While disclosure of minor AI assistance isn’t required, authors are accountable for the authenticity and quality of their own work; all content must undergo a rigorous manual review before being submitted.

You must notify us of any AI-generated text, visuals, or translations included in your manuscript. This information is important for our editorial decision-making process. Failure to disclose the use of generative AI could result in corrective actions post-publication.

Do’s

If you decide to use AI to generate content or images in your submission, you must follow these guidelines.

Be Transparent

As outlined above you must reveal any AI-generated content within your submission. Please specify exactly where this content appears and include this statement with your submission.  

Rigorous Fact-Checking

Do not assume AI output is accurate. Large Language Models (LLMs) are prone to “hallucinations” – producing convincing but entirely fabricated facts, scientific data, or citations. This is especially common when dealing with niche topics or information beyond the model’s training cutoff (which varies by tool). You are responsible for verifying every claim.

Source Integrity

Manually cross-reference all citations. AI often hallucinates “ghost” references that do not exist. Provide a comprehensive list of all resources used, including any data or text originated by an AI.

Correct Citation

AI-generated content must be cited according to APA 6th Edition standard. Example (APA 6th): OpenAI. (2023). ChatGPT (Mar 14 version) [Large language model]. https://chat.openai.com/chat

Avoid plagiarism and copyright infringement:

LLMs could inadvertently reproduce significant portions of existing copyrighted work without attribution. The burden of ensuring the manuscript is free from plagiarism and intellectual property infringement rests entirely on the author.

Authorship and Accountability

AI can’t be listed as an author or co-author. You and your co-authors are 100% accountable for the integrity of the work. By submitting your manuscript, you acknowledge the limitations of these tools, including their potential for bias and knowledge gaps.

Don’ts

Don’t use generative AI to create or modify research data.

Avoid uploading any sensitive, proprietary or personal information into AI platforms. Tools like ChatGPT can log user inputs for model training and/or other business purposes, which could inadvertently expose your intellectual property or confidential data to third parties. Think of these prompts as semi-public records.

Further reading

Committee on Publication Ethics (COPE)’s position statement on Authorship and AI tools.