AI in the Era of Fakes and Deepfakes: Risk of Fabricated Photographs and Identities in Academic Publishing

Abstract

Academic publishing has entered an era of fake, including fake authors who are either real entities using fake credentials, or totally concocted personalities that give the impression of real humans. Both can be achieved via the use of artificial intelligence (AI) and software that is capable of completing such a task, and ultimately a deepfake is created. The creation of fictitious deepfakes, even more so when assisted or driven by AI, allows creators to not only establish a fake image or photo, but also embed it within a fake context (e.g., profile). For whatever reason, there are risk of deepfakes during manuscript submission and the publication process, as well as on academic social network sites, like ResearchGate, but are academics, journals and publishers sufficiently prepared to detect them?.