ChatGPT: Ally or Enemy in the World of Medical Publishing?

The International Committee of Medical Journal Editors (ICMJE) has released new guidelines for researchers wanting to use artificial intelligence (AI) language models for scientific manuscript writing and revisions. The guidelines state that these tools should not be listed as authors and that researchers must note how AI-assisted technologies were used. While AI tools can help with tasks such as writing, analyzing data, and catching mistakes, they are also prone to errors and can generate text that seems plausible but may not be supported by data. Meanwhile, the use of AI tools in medical publishing is here to stay, with experts suggesting that these tools could help limit linguistic disparities in scientific publishing and alleviate the burden of some monotonous or mechanical tasks. However, stakeholders warn that AI models are a black box and that researchers using these tools must be aware of methodology to prevent any incorrect, incomplete, or biased output. The guidelines recommend that authors of manuscripts using AI technology to write should carefully review and edit the results, provide all the prompts used to generate new text or analytical work, and avoid plagiarism. All authors of a manuscript are solely responsible for the content in articles that used AI-assisted technology.