China has implemented comprehensive regulations restricting the production of 'deepfakes' since the 10th of this month.

This is part of measures to prevent anti-regime sentiment, which is linked to the recent blank paper protests. It is interesting in that it is a practical ban measure in contrast to the European Union or the United States, where it remains at the level of recommendations on the grounds of freedom of the press. Of course, China's measure seems possible because of the internet censorship system called the Great Firewall, which has been in operation for over 20 years. However, the burden of compliance with transparency and information disclosure is enormous, so doubts about its internal and external effectiveness are likely to continue to arise.

Deepfake technology has spoken from the beginning about the need to label synthetic content, but it was clear that there were limitations in how to ensure this. Furthermore, as the expression "the internet is forever" reveals, once content is created and spread, it is extremely difficult to completely erase it. Moreover, even after content is deleted, the collective consciousness of viewers does not disappear. This was also confirmed through the case of the 'April Voice' content that criticized the Shanghai lockdown policy, which influenced the blank paper protests.

Historically, technology cannot be stopped from penetrating society. The statistic that 95% of deepfake videos worldwide are pornographic, the fake surrender announcement video of the Ukrainian president in the early days of Russia's invasion, and Bruce Willis's appearance in commercials despite his aphasia, are clear examples of the dangerous reality of deepfake technology that affects society as a whole. However, what we should really be wary of may not be the way this technology is regulated. New ways of manipulating the truth have always existed, and constantly focusing on the latest technology always ends up being a losing game. Instead, what we should focus on is why these productions are created and how they are spread, that is, focusing on the social factors that underpin the spread of false narratives.

"Deepfake technology is morally questionable, but not inherently wrong."

Adrienne de Ruiter, an ethicist and political philosopher, in her research, argued that 'expression of those who have not consented', 'intentionally deceiving the viewer', and 'malicious intent' make the results of this technology immoral. It reveals that the creator and the viewer separated from this technology, that is, human intention, is the subject that should be guarded. In particular, it is difficult to regulate the manifestation of the creator's intention in the case of micro-targeted deepfake content aimed at individuals with relatively strong social influence, such as entertainers and politicians.

So how should future alternatives be created? Two main directions can be suggested.

First, we need to recognize and acknowledge that we exist in a world of cameras and recognition systems. The author of this article, and you reading this article, spend most of our daily lives in front of mobile phones and laptops equipped with cameras.From the perspective of the system, human behavior is material for algorithms.

Cameras that check if a child is being taken care of properly exist for an ideal relationship between parents and caregivers, but they are also inhuman subjects who learn and perform the intention to restrict humans from humans. The awareness that we are living with these new subjects can help manage and respond to immoral intentions associated with deepfakes.

Second, community-based education on the subject needs to be created and disseminated. We tend to seek a sense of belonging within weak connections within digital systems. This is a phenomenon related to the lack of a sense of belonging from social groups that was strengthened by the pandemic, where we want to believe that we are connected to someone invisible in terms of tastes or interests. Checking TikTok repeatedly until 2 am, frequent access to Naver Wiki that is not expected to be accurate, continuous story checks on Instagram, neglecting group chat rooms with little interest, etc. are examples of this.

Deepfakes tend to stimulate a sense of belonging from these weak connections, but because there is no deep interest in the subject, it is relatively easy to dismantle the influence of related content. While it was difficult for individuals to verify the authenticity of deepfake content intended to undermine the credibility of politicians, the results of one project where a political party was able to distinguish falsehoods prove that an education program based on the community's perspective, values, and practices can be an effective alternative. This means that platform service companies where deepfake content is shared can create strategic opportunities by establishing and proposing unique community responses to users.

While there are positive examples of the use of deepfake technology, such as the movie 'Fast and Furious', which brought back Paul Walker, who died in an accident, by synthesizing his face over his younger brother's acting, real-world examples of one woman's life being ruined by a fake sex video targeting a female journalist are also occurring.

It is necessary to remember that actors in the film industry are currently the most protected people from deepfake technology. When the target becomes an ordinary person, it seems that current society cannot yet answer what kind of response is possible. Before expecting legal regulations, perhaps the most necessary first step is self-awareness directed towards ourselves who check deepfake-related content on social channels such as TikTok as a form of entertainment.

*This article is based on the original content published on February 14, 2023 inElectronic Times named column .

References

China is about to get tougher on deepfakes in an unprecedented way. Here's what the rules mean

Bruce Willis in commercials… retired and deceased actors also appear in new works 'Deepfake controversy'

The Distinct Wrong of Deepfakes

A.I. Among Us: Agency in a World of Cameras and Recognition Systems

Using deepfakes for experiments in the social sciences - A pilot study⁠⁠⁠⁠⁠⁠⁠