The impacts of deepfakes on journalism: Risks, opportunities, and changes in news organizations

Abstract Deep learning technology allows for the creation of synthetic audio and video content that appears authentic, offering both risks and opportunities for journalism. Such synthetic content may enhance personalization, visualization, and immersion, but challenge the detection of disinformation. To date, there is limited empirical knowledge about the impacts of deepfakes on journalism. Thus, this study’s goal is to explore how journalists assess potential risks and opportunities and how news organizations respond to deepfakes or synthetic content. Interviews with representatives of major Swiss news organizations show that journalists predominantly consider deepfakes as a form of disinformation, and that they have trouble imagining the technology’s potential opportunities for journalism. Moreover, although specialized technological knowledge in news organizations is crucial, mitigating the risks posed by deepfakes and taking advantage of the opportunities synthetic content also require a reorientation toward professional journalistic norms and raising awareness among journalists and audiences alike. These results highlight that deepfakes make it necessary for journalists to renegotiate the role of journalism, professional norms, and the boundaries of AI manipulation.