The Social Impact of Deepfakes

Prior research has largely focused on the potential societal consequences of AI-generated misinformation rather than its real-world prevalence.

“However, prior research has largely focused on its potential societal consequences rather than its real-world prevalence.”

AI-generated content has become increasingly prevalent on social media.

“AIgenerated content (e. g., deepfakes) - artificially generated videos, images, and audio designed to mimic real individuals - has become increasingly prevalent on social media, raising critical concerns about their role in spreading misinformation (Hancock and Bailenson 2021; Groh et al. 2024; Vaccari and Chadwick 2020).”

The characteristics of AI-generated misinformation on social media are poorly understood.

“Despite growing concerns, the characteristics of AIgenerated misinformation on social media are poorly understood.”

There is a lack of robust detection techniques for text deepfakes.

“highlighting the lack of robust detection techniques for text deepfakes and the overlap with human-generated fake content.”

Recent advances in technology for hyper-realistic manipulations to video pose a threat to traditional standards of evidence and trust in media.

“Recent advances in technology for algorithmically applying hyper-realistic manipulations to video are simultaneously enabling new forms of interpersonal communication and posing a threat to traditional standards of evidence and trust in media1–8.”

Deepfake videos have the power to modify and implant false memories, altering people's attitudes toward each other without valid reasons.

“Interpersonally, deepfake videos have the power to modify and implant false memories, altering people's attitudes toward each other without valid reasons.”

The UK government announced its intention to criminalize online self-harm content in late 2022 following the Molly Russell case.

“A notable example is the tragic case of Molly Russell, where the UK government announced its intention to criminalize online self-harm content in late 2022 [6].”

Audiovisual content is met with credulity as videos are accepted almost axiomatically as accurate depictions of reality by journalists, politicians, jurisprudence, and everyday citizens.

“Audiovisual content is met with credulity as videos are accepted almost axiomatically as accurate depictions of reality by journalists, politicians, jurisprudence, and everyday citizens (Farish, 2020;Fallis, 2021;Hancock & Bailenson, 2021).”

Deepfakes pose significant problems across various domains, such as misinformation, political manipulation, propaganda, reputational damage, and fraud.

“Deepfakes can be used for entertainment, education, and research; however, they pose a range of significant problems across various domains, such as misinformation, political manipulation, propaganda, reputational damage, and fraud.”