The Dark Side of Artificial Intelligence: The Possible Risk of Falsifying Images for Scientific Articles.
Piero Antonio ZeccaReguzzoni MarcellaBrambilla AndreaProtasoni MarinaBorgese MarinaRaspanti MarioPublished in: Microscopy and microanalysis : the official journal of Microscopy Society of America, Microbeam Analysis Society, Microscopical Society of Canada (2023)
This article explores the potential risks associated with using artificial intelligence (AI)-generated images in the field of microscopy. It discusses the current state-of-the-art AI-based image-generation techniques and their limitations. It investigates the potential risks associated with the illegal use of AI-generated images, including their use in creating falsified scientific data and the consequences of such misuse. The article concludes by exploring possible solutions to mitigate these risks, such as implementing robust authentication methods and developing ethical guidelines for using and disseminating AI-generated images in the field of microscopy. Additionally, the article also presents the results of a survey involving 101 professionals, showing that the recognition of authentic and entirely AI-generated images is performed well. But, the detection of hybrid images could be improved.