Tracer Newsletter #42 (27/01/20)-Satirical fake video of Bernie Sanders goes viral on Twitter, with some commenters appearing to believe it is authentic
Welcome to Tracer, the newsletter tracking the key developments surrounding deepfakes/synthetic media, disinformation, and emerging cybersecurity threats.
Want to receive new editions of Tracer direct to your inbox? Subscribe via email here!
Satirical fake video of Bernie Sanders goes viral on Twitter, with some commenters appearing to believe it is authentic
A satirical fake video that depicts Bernie Sanders quoting an infamous meme went viral on Twitter, with many commenters on the original post appearing to believe that the video was authentic.
What does the video show?
The video appears to have first been posted to Reddit on the 26th December 2019, with the recent viral repost coming from Twitter user Datashade on January 23rd. The video footage is taken from November’s Democratic presidential debate, with the original audio replaced with an impersonator imitating Sanders’ voice quoting an infamous meme. This has then been combined with the crudely edited video footage that attempts to match footage of Sanders’ existing lip movements with the new audio. The video was then posted with an accompanying ironic caption from Datashade, receiving 1.7m views, 120.7k likes, and 31.5k retweets by 27th January.
Crude fake videos continue to deceive online
While the video’s creator clearly intended the video to be understood as satirical and for viewers to recognise the meme being referenced, many commenters appeared to believe the video was real and condemned Sanders for making misogynistic comments. Given the crude nature of the video editing and impersonation, this presents a worrying example of how even very low-quality fakes could be used to spread disinformation. It also raises key questions regarding the degree of realism required for a satirical artist to label their work as fake, and what are the underlying elements that cause a user to fool for such low-quality fake content.
Synthetic media company Pinscreen presents a live faceswapping installation at the World Economic Forum
Pinscreen co-founder Hao Li provided World Economic Forum attendees with an interactive demo of the company’s live faceswapping technology and delivered a talk on deepfakes’ potential for misuse.
How did the installation work?
The installation allowed World Economic Forum (WEF) attendees to experience Pinscreen’s realtime faceswapping technology (also known as live facial reenactment) by simply sitting on a chair in front of the camera and selecting from a range of celebrity faces to swap with their own. These celebrity faces were then synthetically mapped in live video of the attendees’ faces, with the result being displayed on a screen opposite. While the results are not entirely photo-realistic, creator Hao Li explained that the installation was designed to raise awareness of deepfakes and get attendees thinking about their potential for misuse.
The risks associated with commodifying deepfake technology
The increasing number of companies developing tools for generating photorealistic “deepfakes” raises serious questions about how these companies can ensure the responsible use of their technology. Li emphasises that Pinscreen’s technology will only be sold to other businesses, although he acknowledges that similar technology is likely to fall into the wrong hands, posing a significant threat to democratic processes. As the tools for creating deepfakes and synthetic media continue to be commodified by companies and open source developers alike, this potential for misuse is likely to continue being the topic of intense scrutiny.
This week’s developments
1) Joe Biden wrongly accused Bernie Sanders of sharing a doctored video of him supporting a call for privatised social security, with Sanders instead making the false claim in a campaign email. (Politico)
2) A new report by the Kofi Anan Commission on Elections and Democracy in the Digital Age found that social media has caused irrevocable damage to global election integrity. (Wired)
3) Disinformation surrounding the Coronavirus outbreak has been circulating on social media, with Buzzfeed News providing a running list of the different forms they’ve encountered. (Buzzfeed News)
4) Twitter announced it will begin labelling world leaders’ tweets with disclaimers if they contain content that violates Twitter’s community standards but are still deemed to be in the “public interest”. (CNN)
5) A team of researchers from the University of York synthetically recreated a short sample of a 3,000 year old Egyptian mummy’s voice using CT scans, 3d printing, and a voice synthesizer. (Gizmodo)
Opinions and analysis
How do people decide whether to trust a photo on social media?
Emily Saltz from the New York Times’ News Provenance Project shares the key findings from research exploring the factors people consider when judging the credibility of photos in news media.
Media forensics and deepfakes: An overview
University Federico II of Naples’ Luisa Verdoliva reviews old and new techniques for analysing digital visual media and detecting deepfakes, with a focus on limitations and emerging challenges.
How misinformation overwhelmed our democracy
Sean Illing outlines how politicians’ cultivation of “manufactured nihilism” and the calculated use of “misinformation zone flooding” has resulted in an “epistemic crisis” for US democracy.
How to exploit satire to become a popular source of false news
Daniel Funke analyses the background of a Facebook disinformation network that attracted a significant readership by reappropriating satirical articles and presented them as authentic news stories.
Want to receive new editions of Tracer direct to your inbox? Subscribe via email here!
Working on something interesting in the Tracer space? Let us know at info@deeptracelabs.com
To learn more about Deeptrace’s technology and research, check out our website
