Can you believe what you see? Unraveling the illusion of reality in the age of AI

The Korea Times

The Korea Times

Opinion

Chyung Eun-ju and Joel Cho

Can you believe what you see? Unraveling the illusion of reality in the age of AI

Posted : 2023-08-01 16:30

Updated : 2023-08-01 16:30

Chyung Eun-ju

Joel Cho

By Chyung Eun-ju and Joel Cho

Earlier this year, a photo of Pope Francis wearing a trendy white puffer coat went viral, only to be uncovered as an AI-generated image. From that point on, the discussion of the ethics revolving around the use of AI to generate images of public figures arose, especially considering that thousands of people initially believed it to be true.

Recently this matter has taken on another level of significance with the use of AI technology to generate the likeliness of public figures, coincidentally, reminiscent of the technology explored in the first episode of the latest season of the Netflix series "Black Mirror."

Volkswagen has been in the middle of the recent discussion revolving around the ethics of using AI-generated likenesses of public figures, with the release of an ad campaign in Brazil that used AI technology to "recreate" the famous singer Elis Regina, who passed away in 1982, singing along with her daughter Maria Rita.

This matter sparked a controversial debate on whether such use of AI technology to "revive" a person crosses ethical lines. On one side of the discussion, we have defenders of the technology, especially since the campaign targeted a very emotional bond the viewers had with the beloved singer, while on the other more skeptical side we had people raising concerns about the ethics of using her likeness four decades after her death.

The controversial ad has led CONAR (the Brazilian National Council for Advertising Self-regulation) to open an ethics action against Volkswagen.

One of the main concerns raised by critics is the ethical treatment of deceased individuals, especially when they are not able to express their consent and desires regarding the use of AI technology to recreate their likeness. When AI is used to create a likeness or revive the appearance of someone who has passed away, it can generally evoke strong emotional responses and memories from the public. Supporters of this technology argue that it can be a powerful way to reconnect people with beloved figures from the past, tapping into the nostalgia and emotional bond viewers had with the deceased, whether this be a family member or a public figure. However, this aspect also poses a clear ethical dilemma, as it may raise questions about exploiting the memory and legacy of the deceased individual for commercial or promotional purposes without their consent. How can we presume the desires and consent of an individual that is no longer among us to do so?

Deepbrain AI, a South Korean company, has grabbed the spotlight with their service Re;memory, which offers a solution for those who are grieving the loss of a loved one. Through AI and a few photos of the deceased, Re;Memory provides a way to talk to the virtual representation of the deceased, potentially offering closure; a way to have one last conversation. Reviving the deceased changes the social norms regarding grief and human-AI relationships. Will society accept these virtual interactions as a natural part of the grieving process?

In the case of the late singer Elis Regina, she might not have explicitly consented to her images being used in this manner. This lack of consent infringes upon their personal autonomy and the right to control how their likeness is portrayed in the public domain, especially in the context of a commercial activity such as an ad campaign.

Another key issue comes with the fact that AI is blurring the lines between reality and fiction. AI-generated content has the potential to deceive and mislead the public, as seen in the case of the Pope Francis photo, where so many people actually believed the viral photo to be real. When such content is indistinguishable from reality, it can lead to the spread of misinformation and harm the credibility of authentic information, especially considering that there are so many fake news stories circulating on social media.

Tom Graham, the CEO of Metaphysics, a generative AI startup, known for making a Tom Cruise deepfake, had said to the Wall Street Journal, "In the next couple of years, it will be possible for regular people to create a photorealistic version of somebody else. Graham suggests that having clearer ownership standards, such as copyright, would enable both celebrities and ordinary individuals to more readily submit requests to remove unauthorized usage of their likenesses or take legal measures against such actions.

AI technology is a terrifying new world where we and our children cannot fully trust what we see with our own eyes. There is a need to discuss the issue at hand to provide more direction in order to balance innovation with responsible use.

Chyung Eun-ju (ejchyung@snu.ac.kr) is studying for a master's degree in marketing at Seoul National University. Her research focuses on digital assets and the metaverse. Joel Cho (joelywcho@gmail.com) is a practicing lawyer specializing in IP and digital law.