Ethical Guidelines for Using AI in Academic Writing

Ethical Guidelines for Using AI in Academic Writing
   1. Manuscripts prepared with use of generative AI in all steps must be done with caution. The AI generated contents must undergo thorough review and careful editing and are recommended to explicitly disclose in the manuscripts during the peer review process, and will appear in the published version. Full responsibility for the accuracy, integrity, and originality of the AI generated contents belongs entirely to all authors.
   2. Authorship and Accountability: Authors are personally accountable for the accuracy, originality, and validity of all content, including any material generated or assisted by AI tools. AI and AI-assisted technologies do not meet the criteria for authorship.
   3. Permitted Use: Authors may use AI to improve language clarity and readability. Using AI for minor language editing, grammar correction, or structuring is permitted, provided the output is overseen and heavily edited by a human.
   4. Prohibited Use: Relying on AI to generate core ideas, literature reviews, or to fabricate and interpret research findings is strictly prohibited. Citing AI-generated text as a primary source is also prohibited. AI tools cannot be credited as authors because they lack accountability for the integrity, originality, and validity of research work. AI tools must not be listed as authors.
   5. Data Privacy and Security: Authors must ensure that patient data, confidential peer-review information, or unpublished manuscripts are never uploaded to open-source AI tools, as this may violate patient privacy and intellectual property rights.
   6. Eproducibility: Authors must provide sufficient documentation, artifacts, or non-AI baselines to enable the independent replication of any AI-assisted research step.
   7. Mandatory Disclosure: Authors must disclose all AI usage even for minor language editing or writing polish within the dedicated manuscript section.
   8. Consequences: The CSDJ journal reserves the right to issue a correction, erratum, or retraction if undisclosed or non-compliant AI usage or integrity gaps are discovered.