In the rapidly evolving landscape of academic publishing, generative AI is redefining the boundaries of authorship and integrity. As scholars, how do we harness this powerful tool while safeguarding the ethical standards of our work?
Recent guidelines from multiple authoritative bodies have established that while AI tools like ChatGPT can enhance research and writing, they do not meet the criteria for authorship. Key publishing and ethics bodies, including The Committee on Publication Ethics (COPE),
爱思唯尔
,
Springer Nature
,
Science Magazine
,
Taylor & Francis Group
,
MDPI
,
Wiley
and
Sage
emphasize that AI tools must not be listed as authors and that their contributions should be transparently disclosed.
These guidelines are critical as they help maintain the integrity of scholarly publications by ensuring clear accountability and preventing the spread of misinformation potentially generated by AI tools.
Here's a summary (with weblinks) that includes some publishers, along with highlighting the specific guidelines related to citing generative AI tools in academic writing:
- ??????: AI cannot be listed as an author; transparency in AI use is mandatory | ???????? ???????????????? ???? ???????? ?????????? ???? ?????? ?????????????????? ?????? ?????????????? ???? ?????????????? ??????????????
- ???????: AI tools cannot meet authorship criteria | ???????????????? ???? ?????????????????????????? ??????????????????????????, ?????????????????? ?????? ???????????? ?????? ???????????? ???? ???? ??????????????????????
- ????????????????: AI contributions must be transparently declared | ?????????????? ???????????? ?????????????? ???????????????? ?????? ???? ?????? ???????? ???? ?????? ???????????????? ?????? ?????????????? ??????????????
- ??????????? ???????????????????: AI cannot qualify for authorship; requires transparency and disclosure of AI's role | ???????????????? ???????????????????? ???? ???? ?????? ???? ?????? ??????????'?? ?????????????????????? ?????????????? ???? ????????????????
- ???????? ???????????????????: AI tools are not eligible for authorship; emphasis on transparency | ?????????????? ???????? ?????????????? ???? ?????????????????????? ???? ?????????? ???????????????????? ?????? ?????????????? ?????? ?????????????????????????? ???? ?????? ????????????????
- ?????????????? ????????????????: Similar to others, no authorship for AI; insists on clear acknowledgement of AI usage | ???????????????? ???????????????? ???????????????????? ???? ?????? ????-?????????????????? ?????????????? ?????? ?????? ???????????? ???? ?????? ?????? ???? ???????????????? ?????? ???????????????????? ??????????????????????
- ???????????? ?????? ??????????????: AI cannot be authors; must be transparent about AI help | ???????????????????? ???????? ?????? ?????????????? ???? ????'?? ????????, ???????????????????? ???????????????? ???????????????????? ???? ?????????????????? ???? ???? ??????????
- ??????????: Prohibits AI as authors; mandates clarity in AI contributions | ?????????????? ???????? ???? ???????????????? ????'?? ???????? ????????????????????, ?????????????????? ???????????????? ?????????????????????????? ?????? ?????????? ?????????????????? ???? ????
This overview informs researchers and academics of the current standards and serves as a foundation for dialogue on AI's evolving roles in research. Engaging with these insights could greatly enrich the ongoing discussions in academic circles about AI tools' potential and limitations in scholarly work.
Temporary Lecturer at Rajarata University of Sri Lanka
7 个月I think both IEEE and ACM says the same. My question is what happens when we find puplications that has whole sections (parahraphs or multiple paragraphs) generated using AI? Guidelines usually say that we need to acknowledge the use of AI. But usually the current GAIs are not that precise (or specific) when describing a something. So if most of the publication (or thesis or dissertation for that matter) is just generated text, that can not be "that good". I mean can GAI generate things like implementation, testing or results analysis in a research? What is your opinion about it sir?