The application of artificial intelligence in scientific articles has recently caught the attention of the media. The ethical application of this breakthrough is one of the issues that engineers should examine in their scientific and operational work today.
A professor recently got a paper for peer review. While reading this article, he came across an unusual phrase: “Regenerate Response.” This phrase had nothing to do with the rest of the paper, but the professor recognized it as one of ChatGPT’s AI features. Based on this, the article reviewer suspects that the author used ChatGPT artificial intelligence to compose at least some of the article’s text. He believes that when the author copied the appropriate text from the ChatGPT website into his own writing, he inadvertently and unintentionally copied and pasted that weird phrase. Another issue is that the author submitted it to the journal website without a thorough evaluation.
The reviewer discusses the matter with the journal editor and requests the author to clarify it. The author admits to using AI to improve article readability and categorization. To verify this, the editor will request prior versions of the article or other proof of the author’s assertion. However, because the author’s response raises major concerns about the overall validity of the manuscript, the journal rejects it.
Did the article violate ASCE journal guidelines?
One of the most exciting aspects of generative artificial intelligence is its capacity to produce content needed in engineering work. Drafting the correspondence of an engineering office, summarizing financial information, simplifying searches, or other applications that these artificial intelligence engines have can increase the productivity and efficiency of engineering work. Unfortunately, this also has a dark side that can threaten professional ethics, especially when engineers delegate tasks that are part of their professional responsibilities and require engineering judgment to AI.
Generative AI is an excellent tool for creating and presenting content, but like any other tool, its users must use it properly and be aware of its limitations. First and foremost, the data these engines are trained with may be flawed. On the other hand, the commands given to these tools to generate information may be vague and problematic, which causes artificial intelligence systems like ChatGPT to generate false, misleading, and even completely fake content. For example, in one famous case, a lawyer was disbarred for creating a fake court order, or in another case, a biologist published an article, and another person, while reading that article, saw his name among the sources of that article. Articles that did not exist externally.
The ASCE Code of Ethics states that engineers should express their professional opinions honestly and only when they have sufficient knowledge about them. Another part of this regulation states that engineers should only approve and sign the work done by themselves or under their responsibility. In this case, if a person presents an article created by artificial intelligence without first reviewing it and ensuring its content’s accuracy and correctness, that person has violated the above-mentioned ethical standards.
Furthermore, because the content generated by artificial intelligence is based on similar existing content, the generated information may violate copyright and contain plagiarism. In truth, any content generated by artificial intelligence is plagiarism because it is copied from other sources and presented as someone else’s work. In other words, this issue is evaluated through the lens of professional ethics because the individual falsely claims to be the creator of the document. For example, a student can be motivated to write an AI-generated text, or a researcher can improve his résumé with an AI-generated article.
According to ASCE guidelines, engineers should receive credit and points only for work they have done themselves or for providing documentation for collaboration on other people’s work. If one’s false claims give that person an advantage over others, this can lead to the article being rejected for unfair competition.
In the example outlined at the beginning of this article, the researcher in issue provided an artificial intelligence-generated article and planned to benefit personally by accepting and publishing it. As a result, his labor offered him an unfair advantage over those who had completed all of the paperwork themselves.
Another point is whether the article’s author mentioned the limited and appropriate use of artificial intelligence in his article. So, because he should have informed the journal about its use in advance, it has created another ethical ambiguity. Considering the limitations and issues mentioned earlier about generative artificial intelligence, sharing that the article was assisted by artificial intelligence would have made the journal’s reviewers review the article more carefully. Considering that the article’s author had hidden this matter and the referee was not aware of it, the article’s author has violated another rule of the journal.
Given the benefits and drawbacks of applying generative artificial intelligence, several publishers have altered their guidelines and regulations for properly incorporating these services into scientific works. For example, the ASCE Journal does not examine or publish AI-written articles. If people utilize artificial intelligence in other ways in their work, they should disclose it and be prepared to answer the editor’s subsequent inquiries regarding the significance and specifics of that use.
Reference: ASCE
0 Comments