Is the Danger of Deep Fakes Fake News?

As artificial intelligence continues to advance alongside our other technological capabilities, some dangers have come to light. With all the discussion surrounding the current presidency and its relationship with the media and "fake news", AI and its ability to produce what are known as "Deep Fakes" have been cause for concern. According to the Wikipedia page on deep fakes, they are described as "synthetic media in which a person in an existing image, video, or audio is replaced with someone else's likeness." Using deep learning, these programs study photographs and videos of an individual in order to learn their typical facial movements and speech patterns in order to produce an artificial version of them.

Of course, with many of these new advancements, they start with good intentions. Deep fakes can be used for lighthearted fun like seeing a celebrity sing a funny song, or seeing what it might be like for one of your favorite actors play parts they were not cast in. In the above photo, which is a screenshot of a deep fake program, you can see Jim Carrey's face on Jack Nicholson's character, Jack Torrance, from "The Shining". You can also see a full video here, https://www.youtube.com/watch?time_continue=30&v=8OJnkJqkyio&feature=emb_logo which is a deep fake version of "Back to the Future" with Tom Holland playing Marty Mcfly and Robert Downey Jr. playing Doc. The issues arise when this ability gets into the wrong hands. These deep fakes can also be used to spread manipulative and false information. There are already a few real life examples of how these applications can be used to manipulate its audience. Deep fake videos are being created of political and company leaders from around the world. In June of 2019, congress began an investigation of deep fakes after a concern was raised about their potential effect on elections. A video of Speaker of the House, Nancy Pelosi, which was a cheap fake (a simply doctored version of an existing video) rather than a deep fake, created a comotion on social media which led to the discussion in congress. As stated by the CNN article, no deep fakes of US politicians have been used to spread false information with malintent or create a deceptive campaign, however this does not mean that the capabilities are not there. In fact, it has already been used in the Legislative Assembly Elections in Delhi. Candidate, Manoj Tiwari and his campaign team used deep fakes to produce multiple versions of his campaign video to distribute to different audiences. He created the original video, and then used deep learning to create two fake videos. One of them in english, and the other in the Hindi language, Haryanvi, in order to sway the local Haryanvi speakers to vote for him instead of the rival candidate even though he was not a native speaker. The possibility is there for these incredible technologies to be used for the wrong intent. We must decide now on whether or not these technologies should continue to be advanced and how to regulate them, as well as protect the public from this source of "fake news." Do you believe that deep fakes provide a greater risk than benefit to society?