AI to deceive viewers to appear more and more online. This phenomenon will cause serious consequences if used in the wrong way. In December 2017, deepfake became a noun when the user named “deepfakes” published a series of porn videos on Reddit forum. This user has used AI to transplant famous actors like Scarlett Johansson or Gal Gadot into pornographic actors. Videos with this very similar face immediately caused a wave of protests. Photo: jta. In 2018, Reddit has closed a series of deepfake video sharing subreddit. Twitter then joined in when blocking shared posts. An adult movie sharing website also prohibited these types of videos. Photo: Reddit. With deepfake, input data is available pornographic videos and images of actors available online, using open source algorithms of TensorFlow or Keras to process. After a period of “learning”, the machine was able to pair faces with high similarity. Photo: Allure. Videos using deepfake technology have caused moral concerns when false information spreads on social networks. Photo: Techhnews. In January 2018, the application called FakeApp launched. This application supports users to make deepfake videos more quickly on computers. This software has an easy-to-use interface and simple usage, with 3 basic commands. In the tutorial article, the creator of this tool sends a message “if you use it, use it responsibly”. Photo: AP. In April 2018, BuzzFeed posted a face-to-face video of former US President Barack Obama with the voice of others. This video is made by a single person via the FakeApp application. He said it took 56 hours to summarize Obama’s face data. Face-transplant videos of powerful people are easily spread and deceive millions of users on social networks. Photo: BuzzFeed. Deepfake fake the speech of former President Obama The video shows the scaryness of deepfake technology when disguising the speech of former President Barack Obama. Last month, the video showed that US House of Representatives President Nancy Pelosi, 79, appeared to be drunk when speaking at an event. The video quickly spread on social networks, broadcast on television and shared by US President Donald Trump. Photo: Crikey. On June 9, a video with fake images of Mark Zuckerberg and Facebook CEO was posted on Instagram. In the video, the image of Mark Zuckerberg is sitting in his office and talking about Facebook’s brutal power. Photo: Washington Post. Video with a description with “imagine this: A man with control of stolen data of billions of users. All secrets, their lives, their future”, “fake Mark Zuckerberg” saying what Mark will never say in the video. Photo: Vanity Fair. On June 13, during a hearing before the US House of Representatives on deepfake, Democratic Party representative Adam Schiff warned that the technology would have serious consequences for the US presidential election in the year of 2020. “The fake videos made by AI are a threat to national security, we need to fight it”, said Adam Schiff. Photo: Reuters. Two weeks ago, a video posted on YouTube showed that Game of Thrones Kit Harington actor, in the role of Jon Snow, apologized for controversial content in the last season of the movie. But in fact, this is a product of deepfake technology and that Harington never uttered those words. Photo: Fortune. In the future, deepfake will definitely become personalized, target specific people and not just stop at video. The fake stories will be produced to distract more and more people. Photo: Partkito. When the news is highly personalized, it will become much more difficult to monitor and prevent. This is when it becomes especially scary. At that time, the ability of individuals to think critically will be a big problem in the period of deepfake development. Photo: USA today.