Recent advances in deep generative models have enabled the generation of realistic synthetic media or deepfakes, including synthetic images, videos, and text. However, synthetic media can be misused for malicious purposes and damage users' trust in online content. This dissertation aims to address several key challenges in defending against the misuse of synthetic media. Key contributions of this dissertation include the following: (1) Understanding challenges with the real-world applicability of existing synthetic media defenses. We curate synthetic videos and text from the wild, i.e., the Internet community, and assess the effectiveness of state-of-the-art defenses on synthetic content in the wild. In addition, we propose practical low-cost adversarial attacks, and systematically measure the adversarial robustness of existing defenses. Our findings reveal that most defenses show significant degradation in performance under real-world detection scenarios, which leads to the second thread of my work: (2) Building detection schemes with improved generalization performance and robustness for synthetic content. Most existing synthetic image detection schemes are highly content-specific, e.g., designed for only human faces, thus limiting their applicability. I propose an unsupervised content-agnostic detection scheme called NoiseScope, which does not require a priori access to synthetic images and is applicable to a wide variety of generative models, i.e., GANs. NoiseScope is also resilient against a range of countermeasures conducted by a knowledgeable attacker. For the text modality, our study reveals that state-of-the-art defenses that mine sequential patterns in the text using Transformer models are vulnerable to simple evasion schemes. We conduct further exploration towards enhancing the robustness of synthetic text detection by leveraging semantic features. / Doctor of Philosophy / Recent advances in deep generative models have enabled the generation of realistic synthetic media or deepfakes, including synthetic images, videos, and text. However, synthetic media can be misused for malicious purposes and damage users' trust in online content. This dissertation aims to address several key challenges in defending against the misuse of synthetic media. Key contributions of this dissertation include the following: (1) Understanding challenges with the real-world applicability of existing synthetic media defenses. We curate synthetic videos and text from the Internet community, and assess the effectiveness of state-of-the-art defenses on the collected datasets. In addition, we systematically measure the robustness of existing defenses by designing practical low-cost attacks, such as changing the configuration of generative models. Our findings reveal that most defenses show significant degradation in performance under real-world detection scenarios, which leads to the second thread of my work: (2) Building detection schemes with improved generalization performance and robustness for synthetic content. Many existing synthetic image detection schemes make decisions by looking for anomalous patterns in a specific type of high-level content, e.g., human faces, thus limiting their applicability. I propose a blind content-agnostic detection scheme called NoiseScope, which does not require synthetic images for training, and is applicable to a wide variety of generative models. For the text modality, our study reveals that state-of-the-art defenses that mine sequential patterns in the text using Transformer models are not robust against simple attacks. We conduct further exploration towards enhancing the robustness of synthetic text detection by leveraging semantic features.

he specter of deepfakes and artificial intelligence enabled media productioncontinues to exacerbate the fear brought on by a degraded ability to discern the real from the fake, syn- thetic, or fabricated in a networked society. While these fears are well-founded especially as they pertain to issues of involuntary pornography their introduction into an already oversaturated media landscape, if anything, extended trends in mediated indeterminacy already being fostered by the universalization of social media platforms. Sites such as Facebook, Instagram, Snapchat and TikTok, made more explicit the contingency and per- formative nature of identity. As younger generations came of age through social media they learned how to navigate and present themselves through it in novel ways unique to each platform. Oftentimes, these strategies were harmful to people’s perception of themselves and their mental health. Other times, however, it gave them the ability to experiment with new forms of identity more in line with how they actually felt. Further, more experimentation through ubiquitous mediation extended what kinds of identities are possible in general as well. In turn, the discovery and extension of identity has led to the evolution of desire. Identities and desires hitherto not possible in a physical space precipitated the creation of new objects of desire that can be pursued and materially experienced regardless of their virtual nature. Deepfakes, and now generative AI, anticipate a further, exponentially more complicated relationship with identity and desire formation through the adoption of increasingly unreal presentations of each. / Media Studies & Production

La mise au point de milieux de culture permettant de mettre en évidence rapidement et de manière reproductible des micro-organismes exigeants représenterait une évolution significative dans le contrôle des produits et procédés industriels. Deux milieux de culture synthétiques ont été élaborés pour répondre à ce besoin. Le développement des milieux de culture a été réalisé en combinant l’analyse de composés nutritionnels à l’étude de certains métabolismes bactériens. Les formulations des milieux de culture obtenues permettent ainsi la croissance de micro-organismes aux exigences nutritionnelles très variées comme ceux fréquemment isolés dans l’eau industrielle. En parallèle, des expériences sur des biofilms multi-espèces, utilisant des bactéries isolées de systèmes d’eau industrielle, ont permis d’observer que la survie des bactéries en milieu pauvre en nutriments, est dépendante d’interactions coopératives, basées sur le « swarming » et le « hitchhiking ». / The development of culture media to quickly and reproducibly detect fastidious microorganisms represents a significant change in the control of industrial products and processes.Two synthetic culture media have been developed to reach specifications. The development was performed by combining the analysis of nutritional compounds to the study of specific bacterial metabolisms. The formulations of the resulting culture media allow the growth of microorganisms with different nutritional requirements as those frequently isolated in industrial water.In parallel experiments on multi-species biofilms, using bacteria isolated from industrial water systems, it was observed that the survival of bacteria in nutrient-poor environments is dependent on cooperative interactions, based on "swarming" and "hitchhiking".

Introduktion: Syntetisk media omfattar teknik som berör många områden, t ex syntetisk text- och röstbearbetning, videomanipulation, bildmanipulation, digitala människor och ansikten. Under de senaste åren har den tekniska utvecklingen accelererat ytterligare, till en informations- och kommunikationsteknologi (IKT) som omfattar möjligheten att använda syntetisk media alltmer. Detta har lett till implementering av nya informations- och kommunikationsverktyg för hälsokonsumenten, bl. a med förhoppning om ökad individuell empowerment. Det är således intressant att undersöka hur syntetisk media beskrivs vara möjligt att användas hälsofrämjande inom folkhälsoarbete samt beskriva hur det påverkar empowerment hos hälsokonsumenter. Syfte: Syftet för detta examensarbete har varit att redogöra för vilka syntetisk media som används idag som informations- och kommunikationsverktyg inom hälsopromotion och prevention samt undersöka deras konsekvenser på hälsokonsumentens empowerment. Metod: En strukturerad litteraturöversikt med tematisk analys enligt Braun &. Clarks (2006) modell. Resultat: Totalt har 14 vetenskapliga artiklar ingått som urval i den tematiska analysen. Efter tematisering har slutligen 5 teman kunnat presenteras; Hälsokonsumentens karakteristika, Målområden för digital hälsa, Digitalt verktyg-syntetisk media, Intervention genom digitalt verktyg-syntetisk media, Konsekvenser för hälsokonsumenten. Slutsats: Syntetisk media kan verka som ett digitalt hälsofrämjande informations- och kommunikationsverktyg och har möjlighet att genom olika digitala interventioner rikta sig till hälsokonsumenter med potential för att stärka deras empowerment. Hälsokonsumenterna har beskrivits som en heterogen grupp, med olika åldrar, kön och bakgrund. I dagsläget är det till arbetets syfte relaterade urvalet av studier dock begränsat. Det finns därmed ett behov av fortsatta studier som följer utvecklingen av syntetisk media i relation till folkhälsoarbete beträffande utformning, effekt, hållbarhet och säkerhet. / Introduction: Synthetic media includes technologies such as synthetic text and voice processing, video manipulation, image manipulation, digital people and faces. In recent years, technological development has accelerated further, to an information and communication technology (ICT) that increasingly embraces the possibility of using synthetic media. This has led to the implementation of new tools for the health consumer, in the hope of increased individual empowerment. It is therefore interesting to investigate synthetic media in the context of public health work and empowerment in health consumers. Aim: The aim of this thesis has been to account for which synthetic media is today used as information and communication tools in health promotion and disease prevention and to investigate their consequences on the empowerment of the health consumer. Methods: A structured literature review with thematic analysis according to Braun & Clark's (2006) model. Results: Totally 14 scientific articles have been selected for the thematic analysis. All together 5 themes could be presented: Characteristics of the health consumer, Target areas for digital health, Digital tools-synthetic media, Intervention through digital tools-synthetic media, Consequences for the health consumer. Conclusion: Synthetic media can act as a digital health promoting information and communication tool and target health consumers through various digital interventions with the potential to strengthen their empowerment. Health consumers have been described as a heterogeneous group. At present, there are limited studies available related to the study aim. There is therefore a need for further studies of synthetic media in relation to public health work related to design, efficacy, sustainability and safety.

This master thesis explores the media representation of deepfakes. The first part summarizes previous research followed by a comprehensive review of deepfakes, including the technology allowing for their emergence, current uses and methods of regulation and detection. The second part connects the phenomenon with important theoretical concepts such as social construction of reality and the crucial role of media in this process. The empirical part consists of research combining two methods - quantitative content analysis and qualitative critical discourse analysis. The research analysis is focused on media articles dealing with deepfakes in order to find out how the media represent this phenomenon. The results show that current media discourse of deepfakes is strongly negative as the media frame them as a security threat. This negative representation is highly speculative since journalists often invent their own stories of future disastrous consequences of the technology for national security due to lack of current examples. The findings show an apparent hierarchy of the harms posed by deepfakes which is present in media coverage, and reflects gender sereotypes and inequality in the current society. Harm in the form of non-consensual fake pornography targeting women is neglected in the media...