Use of Artificial Intelligence (AI)
- Can AI be listed as an author?
- When must I disclose the use of AI?
- Does AI disclosure create a bias in the peer review process?
- What kind of AI use is prohibited?
- How do I disclose AI use?
- Can I use AI to create images or graphs?
- Who is responsible for AI-generated errors?
- What am I expected to check accurately when using an AI tool?
Can AI be listed as an author?
No. AI cannot be listed as an author or co-author because tools are not legal entities capable of taking responsibility for the work.
When must I disclose the use of AI?
Disclosure is necessary when using generative AI to create text, generate figures, check analyses. Routine editing (spelling/grammar, translation) does not requires disclosure. In this case, please write “No AI-based tools have been used in preparing this manuscript".
Does AI disclosure create a bias in the peer review process?
No. In the current scholarly publishing climate, AI is accepted as a valid assistant if used responsibly and appropriately. Nonetheless, researchers must accept accountability when using generative AI to create text, figures, code etc… and trust the review process which requires they disclose how the different tools were used.
What kind of AI use is prohibited?
Authors should never use AI to create, manipulate, or fabricate research data and results.
How do I disclose AI use?
Disclose the specific tool (e.g., ChatGPT, Claude...) and how it was used in the appropriate field during the submission process. Include the name, version, and manufacturer of the tool as well as the prompts you used.
Can I use AI to create images or graphs?
Disclosure is mandatory if AI is used to create or edit data visualizations, illustrations, or images.
Who is responsible for AI-generated errors?
The authors are entirely responsible for the accuracy and integrity of the submitted work, including content assisted by AI.
What am I expected to check accurately when using an AI tool?
You must check the terms and conditions of any AI tool to ensure that the privacy and confidentiality of data and inputs of the material is guaranteed. Particular care should be taken with any personally identifiable data. Images that duplicate or refer to existing copyrighted images, real people, or others’ identifiable products or brands must not be generated. You should always check for factual errors and for any potential bias.
You must also ensure that constraints are not imposed on the use of outputs from the AI Tool in a way that could restrict the subsequent publication of the relevant article or collide with your copyright or the copyright of the journal or publishers.