Abstract

Recently, computer-generated and computer-altered videos known as deepfakes have raised widespread concerns about the harms they may cause to democratic elections, national security, people’s reputation, and people’s autonomy over their words and actions as represented in videos and other media. How can we build towards a critical understanding of not only deepfakes, but also photos, videos, and the role of many other media objects surrounding us that inform us about the world? In this thesis, wanting to take a historical approach and noting the newness of deepfakes, I first investigate a historical case study regarding a manipulated photo from a 1950 U.S. Senate election campaign. Examining hearings conducted by the Senate into the use of misleading media in the election, I investigate how the incident sparked a debate between different groups of people over the trustworthiness of photographs and their proper role in elections. Next, I move forward in time and discuss the nature of deepfakes, presenting a brief history focusing on the different communities—academic, hobbyist, and commercial—that have played a role in the development of different, but related, technologies that all fall under the umbrella term of deepfakes. Some of this history is incorporated into the third part of this thesis, in which I present a teaching module I developed with the goals of guiding students to think critically about photos and videos and of raising awareness about deepfakes.