Introduction
The Benefits of Deep Fake Technology
Deepfake tech has roots in personal and social entertainment. Snapchat incorporated the face swap feature into its collection of filters as early as 2016 (Keogh, 2020). Users eagerly swapped faces with their friends to share with the niche community. Humans have a profound sociological relationship to television media, which Marshal McLuhan studied extensively. Deepfake technology has the potential to expand our ability to reimagine our relationship with digital media on an individual, social, and cultural level.
"Televisual tech shaped a psychological and social environment where the starting point for the personal and political agency was imagination" (Paulos, 2017, p. 84). Television tech has evolved into deeper layers of sharing and imagining within our digital habitats. There is symbolic power in deepfake tech and its ability to shape our environment (Couldry, 2020).
This power has been demonstrated profoundly in history, art, and scientific analysis. With deepfake, tech creators can bring art history to life, create new environments that did not exist before, and change how we interact with media content. An example of this transformative power is using AI with deepfake tech at the Salvador Dalà museum. This interactive experience allows museum guests to see the artist differently, hearing the artist's words. A deeper and richer connection to our cultural world is fostered across space and time, a "presence in absence" (Taras, 2015).
"Those who could best imagine the future, and relay the right images to the world, could best change the world" (Paulos, 2017, p.84).
The Concerns of Deep Fake Technology
(Deeptrace Labs, 2019, p. 1)
Deepfakes emerged on Reddit in 2017 as a way to use artificial intelligence to superimpose celebrities' faces onto pornography (Corry, 2022). Despite the nefarious origins of deepfake technology, we are now distracted by seemingly harmless social media videos and fears of political interference. However, the technology has not strayed far from its roots. As stated by Curran, technology will always be influenced by "the objectives of the people who funded, created, and fashioned it" (2012, p. 36). A 2019 report by Deeptrace Labs found that 96% of deepfake videos involved the creation of fantasy pornography without women's consent (Romano, 2019). In fact, our team downloaded a free app called Face Magic to test deepfake technology, which was recently found to be skirting app store rules and encouraging users to "make deepfake porn in a sec" (Cole, Maiberg, 2022). Another way in which women are harmed by deepfake technology is the website DeepNude which uses machine learning to digitally "undress" photos of women (Deeptrace labs, 2019, p. 8). Couldry's third principle is that media reinforce existing power structures (2020, p. 23). As a new form of media, deepfakes are reinforcing the patriarchal values of our society.
Despite the actual harm and distress this technology causes women, the media focuses on politics and business. According to Deeptrace Labs, "much of the media attention surrounding deepfakes has focused on their potential to undermine democratic processes and enhance cyberattacks against individuals and businesses" (2019, p. 9). While we have seen increased politically motivated deepfakes, they are still relatively quickly flagged and identified. "Awareness of deepfakes alone destabilizes political processes by undermining the perceived objectivity of videos featuring politicians and public figures," said Deeptrace Labs (2019, p. 15).
A new deepfake music video by Pulitzer Prize-winning rapper Kendrick Lamar perfectly sums up the double-edged sword of deepfake technology.
"The line between admiration and degradation is too fine to decipher, much like the line between reality and artificial intelligence in Lamar's video. He has the ability to make deepfakes feel like art, which could help legitimize a technique that encourages revenge porn and political manipulation. Deepfakes have the ability to masquerade as the truth, spreading disinformation in the name of swaying public opinion" (Corry, 2022).
The Battlefield
Digital media policies are explicit in their expectation that users will not impersonate another person on their platforms. However, because digital media platforms are intermediaries, they have the right but not the responsibility to moderate user content (Gow, 2022). Social media platforms like Snapchat, Tik Tok, and Instagram have had significant user engagement with rudimentary deepfake technology through popular filters. Tik Tok may have some of the highest quality deepfake content. For example, some Tik Tok accounts are based entirely on Deep Faking high profiles individuals like Tom Cruise, the Queen of England, and Putin. Regardless of how social media platforms present their official policies, they have been complicit in allowing users to experiment creatively with the technology.
When looking at deepfake content on Google, the media available is disproportionately pornography. A well-known example is actor Kristen Bell discovering that her image had been represented in pornography using deepfake tech. In an interview with Vox (2020), Kristen Bell said, "I was just shocked because this is my face…Belongs to me!... It's hard to think that I'm being exploited." This case demonstrates that women are at risk of sexual exploitation through deepfake technology. There appears to be limited control and moderation of harmful deepfake content. Deepfake tech will undoubtedly play a significant role in what Curran (2012) describes as 'the battle for the soul of the internet.'
Creators go against policies concerning deepfakes; this is evident from their actions. Content ranges from silly videos of Queen Elizabeth dancing to more sinister exploitive content. Under the United States (US) law, internet intermediaries, most of which provide services to Canada, are protected by the Communications Decency Act (CDA) 230, which provides the right and not the responsibility for intermediaries to monitor and manage content that runs through their system (Gow, 2022). Social media platforms and other media websites are complicit in allowing users to experiment with the technology. Policy and lawmakers in Canada and the United States are currently ill-equipped to regulate this new digital technology.
Conclusion
Suggested Readings
If you want to dive deeper into Deepfakes, check out the links below:
Deep Fakes and the Infocalypse : What You Urgently Need To Know
Deep Fakes: The Latest Anti-Woman Weapon - Women's eNews
Deepfakes are a real political threat. For now, though, they're mainly used to degrade women.
Kendrick Lamar's New Deepfake Music Video Is a Problem
References
Cole, S. & Maiberg, E. (2022, May 10). A Popular Face Swap App Is Advertising Deepfakes on Porn Sites. Vice. https://www.vice.com/en/article/epxeae/face-swap-app-on-apple-app-store-google-play-deepfakes
Corry, K. (2022, May 16). Kendrick Lamar's New Deepfake Video Is a Masterpiece. That's Exactly the Problem. Vice. https://www.vice.com/en/article/akvxwa/kendrick-lamar-deepfake-music-video-the-heart-part-5
Couldry, N. (2020). Media: Why it Matters. Cambridge, UK: Polity Press.
Curran, J. (2012). Misunderstanding the Internet. Rethinking internet history. (p. 34-60). Taylor and Francis Group. Routledge
Deeptrace Labs. (2019). The State of Deepfakes. https://regmedia.co.uk/2019/10/08/deepfake_report.pdf
Gow, G. (2022, May 16). Issues in Representation. [Google Slides]. eClass. https://docs.google.com/presentation/d/1EgfbGFzHvDoUaaxUxoB_BaU03Doz6Q3MZsQFXYX0SuQ/edit#slide=id.p
Keogh, B. (2020, March 21). Deepfakes: New Zealand experts on how 'face swap' could turn sinister. Stuff. Retrieved May 23, 2022, from https://www.stuff.co.nz/technology/digital-living/120397261/deepfakes-new-zealand-experts-on-how-face-swap-could-turn-sinister
Poulos, J. (2019). Imagine All the People. The New Atlantis, 58, 84–90. https://www.jstor.org/stable/26609122
Romano, A. (2019, October 7). Deepfakes are a real political threat. For now, though, they're mainly used to degrade women. Vox. https://www.vox.com/2019/10/7/20902215/deepfakes-usage-youtube-2019-deeptrace-research-report
Taras, D. (2015). Digital Mosaic: Media, Power, and Identity in Canada. The New Architecture of Media Power. University of Toronto Press.
The most urgent threat of deepfakes isn't politics. YouTube. (2020, June 8). Retrieved May 25, 2022, from https://youtu.be/hHHCrf2-x6w
You made your own deepfake! This raises the matter of the three forces of the long tail and how making these has become relatively easy for anyone to do. Of course, this raises the further issue that Couldry discusses around sharing and trust. Increasingly, will our awareness of dfs will undermine our trust in what we see and share online? How can this be addressed and what measures can be introduced that allow us to take advantage of the positive-creative potential of dfs versus the negative social implications? The post contains some good research and is well written and an interesting account of this phenomenon.
ReplyDelete