What you see below are photos from an actual video of former United States president Barack Obama, and a DeepFake reenactment of the same video created by Artificial Intelligence.
As Artificial Intelligence grows and becomes more advanced, one of the most concerning worries has emerged: The creation of DeepFakes. These generated videos created by AI can accurately mimic the appearance and voice of real people. Deep fakes have been around for years. but in the growing age of new technology centered around Artificial Intelligence it is becoming significantly more difficult to spot what is real, and what isn’t. The growing threat of AI generated photos and videos poses a significant threat to individuals, society, and the security of the globe. While to most people, current AI photos are easily extinguishable, there are programs and softwares that create accurate and fake renditions of other people.
WHAT ARE DEEP FAKES?
Deep Fakes are photos or videos digitally created using AI technology. They are usually made using GANs (Generative Adversarial Networks.) These programs can take real footage or audio of a person and use it to create fake, however very convincing, content of people. From changing what someone says, to placing someone’s face on someone else’s body. Capabilities such as the ones that have been stated could have the ability to trick, deceive, and cause mischief on an unparalleled scale.
DANGERS POSED BY DEEPFAKES
Many dangers could arise due to AI DeepFakes, dangers on a personal, societal, and global scale. For example on Wednesday August 28th 2024, News reports from South Korea claimed that the country was having a massive “AI DeepFake Emergency”. According to South Korean police, there have been hundreds of messages and videos going around on apps such as telegram that show South Korean women falling victim to DeepFakes. According to South Korea’s president, these videos would use AI to combine the faces of South Korean women to create “adult content”. Reports would say that there would be group chats where photos of women, many of whom were underage classmates or teachers, would be sent their photos turned into sexually explicit images. This is just one possible danger that could cause personal harm to people.
AI deep fakes could also be used to help spread misinformation. Deep fakes could be used to mimic the look and voice of influential people to help spread misinformation about world events or individuals.
RESPONDING TO DEEP FAKES
To help combat the threat of deep fakes, various systems and applications are currently in development that will be able to detect the authenticity of digital content. However there is currently an arms race between individuals who create deep fake content and those trying to make systems trying to detect it. However, with the ongoing development of deep fake technology, it’s becoming increasingly harder to be able to detect this kind of digital content.
Regulating deep fakes presents significant issues. Some places have implemented laws to help prevent their misuse. However the nature of the technology makes enforcing the creation of deep fakes incredibly difficult. This further proves intent and jurisdictional managing issues, complicating legal efforts to control the spread of deek fakes.
CONCLUSION
DeepFakes represent one of many possible dangers of AI, with the potential to cause harm to individuals, society, and threaten global security. We are currently in an age of new and growing technology, and as this technology continues to grow we must remain vigilant. Now the purpose of this article is not to strike fear over AI but to alert people of the possible dangers it could cause despite its many upsides. It’s important to support proper regulations to help negate the potential risks that are posed by Deep fakes created by Artificial Intelligence.