EOTO 2: The Dangers of AI Deepfakes

 

Imagine seeing a video of a world leader declaring war, or your best friend saying something they never actually said. With AI deepfakes, that terrifying possibility is no longer science fiction; it’s already here. When watching old movies and television shows, I always enjoyed seeing how they envisioned the future, but while there have been great advances, there have also been more terrifying ones. Personally, I feel deepfakes fall into the latter category. 

Deepfakes are defined as piece of media in which a picture or video of person is manipulated by ai to make it seem they are saying or do. They date back to as far as the 1990s. During this time, people were making advances in technology to increase the effects they could use in television and movies. CGI Researchers came up with the technology that later created AI Deepfakes, to make more realistic human characters in film. The technology has evolved especially as the technology behind artificial intelligence evolves. The term for deepfakes was coined in 2017 by a Reddit user who had the username "deepfakes". The term mixed the terms "deep learning" which is a type of learning based on machines and artificial intelligence, and "fake", as these depictions are simply not real. 

    Deepfakes present many dangers especially in the age of cancel culture. With cancel culture, many have career ending situations in which they are seen doing or saying something that goes against society's most popular ideas at the time. While this is a dangerous mob mentality mindset to begin with, its dangers are escalated when you throw in complete fiction being confused with reality. There have been many incidents in which deepfakes have been used to mimic celebrity endorsements for a less popular candidate, or even maliciously intended to appear as someone saying defamatory things or expressing ideals that they do not believe in. Other dangerous incidents include non-consensual videos released made to appear to be people who aren't actually there. Many of these types of fictional productions can bring harm to not just celebrities, but also everyday people.

Many may be thinking, "well I am not famous so I can just use the humorous aspect of Deepfakes", and while, yes, you can simply put your friends face on a orange or in a meme, you are still equally subjected to be put in videos so long as your face appears on the internet, even from professional headshot photos. Among those most subjected to being put in videos, women make up 90%. Those most subject to believe them to be true are those apart of the elderly community. Many people are at risk to be put in them, and not just celebrities. 

   There have been continuous efforts for deepfakes to be more restricted and to increase internet saftey. however, this is still a fairly new concept and there have been obstacles in the way of these attempts to put restrictions on this tool. According to an excerpt from the 2023 edition of the Northwestern "Buffet Briefs", several countries began looking into legal measures to regulate deepfake productions and distributions in 2019. There is still no set solution yet as they are currently covered under the 1st amendment as free speech and expression.