Four Guidelines for Avoiding Risks in Academic Ethics
The following specifications must be observed when using AI writing tools:
- Principles of Content Positioning: Always consider AI-generated content as 'smart drafts' rather than final drafts. According to Elsevier publication ethics standards, it is recommended that the proportion of manual revision should be no less than 70%, focusing on the improvement of three parts: research data, personal opinion, and argumentation logic.
- Testing and validation process: After using the platform's built-in 'AI Rate Detection' (based on the GPTZero algorithm), it should be verified a second time using an academic system such as Turnitin. Generally colleges and universities require AI logo content of no more than 15%.
- Reference checking: The 40 system-generated references need to be validated one by one, and it is recommended to 1) check the DOI validity 2) compare the citation contexts for the original intent 3) add 2-3 more recent references (after 2023).
- Declaration and labeling: Explicitly state 'use of AI tools to assist in framework building and first draft writing' in the methods section or acknowledgements section of the paper, see Nature's latest guidelines for AI attribution.
Special note: Direct submission of AI-generated content may be considered academic misconduct, and testing standards are becoming stricter each year at each school.
This answer comes from the articleThousand Pens Writing: an AI tool to assist in completing essay writingThe
May not be reproduced without permission:AI productivity tools " How to Avoid Academic Ethics Risks of AI-Generated Papers?
English 
简体中文 
日本語 
Deutsch 
Português do Brasil