Insider Brief
- Researchers at Northwestern University and the National Institutes of Health warned that generative AI is creating new concerns around “plagiarism of ideas,” where AI systems may reproduce underlying concepts or intellectual contributions without proper attribution.
- In a commentary published in Nature Machine Intelligence, the researchers argued traditional plagiarism policies focused on copied text may be less effective as AI systems increasingly rephrase content while preserving original ideas.
- The authors proposed updating definitions of research misconduct to clarify that users remain responsible for plagiarism and other ethical violations that may occur through the use of AI tools.
Researchers at Northwestern University and the National Institutes of Health suggest that the rise of generative AI should force a rethink of how plagiarism and research misconduct are defined in scientific writing.
In a commentary published in Nature Machine Intelligence, Northwestern’s Mohammad Hosseini and NIH bioethicist David Resnick write that traditional definitions of plagiarism may no longer be enough as AI writing tools become more common in research. According to Northwestern, existing policies largely focus on copied text and theft of ideas, but advances in generative AI make verbatim copying less central because systems can easily rewrite material while preserving the underlying concepts.
The bigger concern, the authors argue, is idea plagiarism — situations where AI systems reproduce concepts, arguments or intellectual contributions without proper attribution. Because that form of intellectual borrowing can be difficult to detect, the researchers propose revising definitions of research misconduct to clarify that people remain responsible for plagiarism committed through AI tools.
The researchers pointed out plagiarism damages the research ecosystem by undermining trust, obscuring the true origins of scholarly work and discouraging original thinking and innovation.
“It is fine and in fact helpful to use GenAI to increase the readability of writing and bounce ideas back and forth, but we know these tools frequently make mistakes of fact and accuracy and have enormous social and environmental impacts,” noted Hosseini, assistant professor of preventive medicine in the division of biostatistics and informatics at Northwestern University Feinberg School of Medicine. “Checking AI output is still the simple and only way to ensure content is correct and reliable.”
Hosseini pointed out that users who rely on generative AI without conducting their own research or closely reviewing outputs may unknowingly introduce plagiarized material. He argued that clarifying definitions of misconduct would reinforce that responsibility for avoiding plagiarism rests with the person using the technology, not the AI system itself.
The authors said the issue extends beyond academia into fields including law, medicine and business as AI systems become more embedded in everyday work.
“Non-researchers should also use GenAI in responsible ways,” Hosseini added. “Plagiarism is an ethical and legal concern not just for researchers but also for students and those working in various professions, such as law, business and medicine.”
