The rapid development of modern science and technology not only makes artificial intelligence technology gradually integrated into our lives but also makes intelligent machines from the dull mechanical movement “alive”. What is artificial intelligence? What we often call artificial intelligence is actually a machine learning behavior through computers to simulate human thinking and behavior.
Professor Nils John Nilsson has defined artificial intelligence as the discipline of knowledge, the science of how to represent knowledge and how to acquire it and use it (2011). According to Professor Winston, artificial intelligence is the study of how to make computers do intelligent work that only human beings could do in the past (2017).
As people’s research on artificial intelligence gets deeper and deeper, the fields covered by artificial intelligence technology are becoming more and more extensive. In addition to the most common and most used AI products in daily life such as intelligent search, automatic planning, and driverless systems, AI technologies such as face recognition technology and AI face changing technology are examples.
While bringing convenience and deep entertainment to humans, it also brings legal and ethical issues such as personal privacy leakage and infringement. In this paper, we will explore the ethical challenges of AI face-swapping technology when applied in the digital age through the benefits and controversies brought by AI face-swapping technology.
What is artificial intelligence face replacement technology?
Artificial intelligence face-swapping technology is also known as face-swapping technology. As the name implies, it means replacing one face with another in an image or video. Artificial intelligence, which has entered the digital era, not only has easy access to massive amounts of data, but its superb computing power and constantly innovative algorithms provide room for the rapid development of artificial intelligence face swapping technology.
Some of the once difficult points such as extracting facial information of the face swap object and transforming the training model can be accomplished by the massive database and AI’s deep learning (Kuang, Guo, Z., Fang, J., Yu, J., Babaguchi, N., & Fan, J., 2021). Simply put, select the object you want to perform face swapping, overlay the selected object, and generate automatic replacement by AI, so that you get the face-swapped photo or video you want.
The advent of DeepFakes technology has brought a remarkable breakthrough in face exchange technology. With the new technology, you only need to have enough photos and a graphics processor to achieve very realistic visual effects, even if you don’t know anything about video editing (Zendran, & Rusiecki, A., 2021).
A typical representative of this is the ZAO app, which was a hit in 2019. In this app, you only need to provide a photo of your face, you can become the main character in major movies and TV dramas, to experience the feeling of being a star.
The development trend of artificial intelligence faces replacement technology
In fact, face exchange technology is not new in the field of film production. This has to mention the most representative Fast and Furious 7. In 2013, Paul Walker was shot in half and unfortunately died in a car accident. His brother replaced him in the scene, through artificial intelligence face exchange technology to complete the remaining part of the shooting, which is also considered to fulfill the dream of the film fans.
The Dali Museum, which opens in 2019, will also bring Salvador Dali back to life through the use of artificial intelligence face-swapping technology, allowing him to interact with people who come to visit. When it comes to bringing the dead back to life, virtual reality technology also opens up more possibilities for artificial intelligence face transplantation.
Artificial intelligence face replacement technology can achieve 3D reproduction through artificial intelligence and virtual reality technology so that people who miss their loved ones can meet their longed-for relatives in the virtual world. In 2020, the Korean MBC TV documentary channel used these two technologies to help a mother “reunite” with her daughter who died three years ago due to illness.
According to the production team, this interactive project took eight months to produce. Technicians through the little girl’s life images, as well as the same-age children to capture the movement. The final synthesis of the little girl’s voice, movements, and facial expressions through artificial intelligence technology.
The use of artificial intelligence technology and virtual reality to create virtual worlds and virtual personalities for families who have lost loved ones should help many families who have lost loved ones to come out of their grief. This seems to be something to look forward to.
But, is artificial intelligence face replacement technology really completely safe and reliable?
Certainly not. There are two sides to any technology, and this technology can be extremely harmful if exploited by the wrong people. Back in December 2017, a user named “DeepFakes” posted a “pornographic video” of Wonder Woman Gal Gadot on Reddit. Although the user’s account was blocked by Reddit under public pressure, the videos were already uploaded to Pornhub by other users.
On Pornhub, we were surprised to find that in addition to European and American movie stars, many famous actresses with superior bodies and looks as well as internet celebrities were “involved” in the shooting of these pornographic videos. Of course, these are fake videos faked by artificial intelligence face-swapping technology, and they may not even know about it. Ironically, however, the DeepFakes technology that brought about the major breakthrough in face-swapping technology is named after the username “DeepFakes” (Banks, 2018).
You may say that it’s because these celebrity groups are highly visible and their photos are very easily available on the internet. We, as the general public, should not have this kind of trouble, right? This is not the case. Just last August, the British media reported such a vicious case of artificial intelligence face swapping.
Alex Woolf, selected by the BBC as Young Arranger of the Year, secretly posted selfies or photos of the daily life of 15 female friends around him to Reddit over a period of two years. After helping him synthesize the women’s faces into the faces of pornographic actresses through artificial intelligence face-swapping technology, he uploaded the fake videos to pornographic websites.
Until early last year, one of the victims received an anonymous email informing her that her photo had appeared on online forums and pornography sites. What Alex Woolf and his accomplices did was only discovered.
Will it be the only obsession of the “actress”?
In addition to the possibility of being faked into pornography and experiencing pornographic retaliation, there are also unscrupulous individuals who use artificial intelligence face-swapping technology to commit economic crimes.
Just at the end of last year, Chinese police uncovered a criminal gang that used “fake faces” for illegal registration and authentication. They use artificial intelligence face-swapping technology to bypass this step of requiring face recognition authentication when registering at relevant government websites.
In the recent hit TV series Under the Skin in China, a man faked a video of his son asking his parents for help through artificial intelligence face-swapping technology. The elderly parents were scammed out of their money and eventually chose to commit suicide together.
In the digital information age, elderly people have always been the “target customers” of fraudulent gangs. Among them, the elderly who do not live with their children and the elderly who do not understand modern technology are the most likely to fall for it.
In the era of digital information technology, the risk of information misuse and leakage of personal biometric features is a major reason why such software frequently raises questions. For groups with high visibility such as celebrities and internet celebrities, the negative impact of moving their faces to some scenes where they should not be is very significant.
As for us, the general public, besides the possibility of being used by people around us, many economic scenarios such as cell phone face unlocking, PayPal, Alipay, etc. currently rely on face online recognition for identity verification. If a pseudo-video is generated by the wrongdoers using the expression manipulation method in deep synthesis, the loss of personal money and reputation caused by this is also unacceptable.
It is worth mentioning that the completion of the initial face-swapping technology requires hundreds of pictures of face-swapping people. And with the continuous improvement of technology in recent years, now only 3-5 relevant photos are needed, and artificial intelligence technology can generate face-swapping photos or videos.
The face is obviously our most important identification symbol, but in this society now, it only takes a few minutes or even seconds to steal these things. Silently, we don’t even know who and where the other person is, and what use they are going to take our photos for.
How should we respond to this ethical challenge?
With the rapid development of modern society, the trend of high technology in the development of science and technology in society has soared. And the development of technology is bound to cause friction between social culture nature and ethical consciousness.
In the digital, intelligent era of big data, human fingerprints, and voices, are turned into data into a member of the Internet, and now this data encapsulation to the entire face. The data leakage potential of big data itself coupled with the emergence of artificial intelligence face-swapping technology, whether one’s own facial bio-information will be easily exploited in this way, has become a reason for criticism and opposition.
Currently, China has enacted the Personal Information Protection Law which plays an important role in protecting the personal information security of citizens. However, after users upload their personal photos to third parties, is the protection system of the third party’s own platform safe? Will it be maliciously breached, thus causing privacy leakage? This still needs our attention.
In addition, in our daily life, we should pay attention to protecting the security of personal privacy. When using applications that require uploading personal information, you should carefully check whether the user privacy requirements and instructions in the user agreement are regulated. Meanwhile, it is also necessary to set friend permissions and screen friend circles appropriately.
More importantly, a young group should communicate more with parents and grandparents in, a timely and popularize the knowledge of artificial intelligence face replacement technology. And explain to them that if they receive a help video do not care and are confused, should be the first time to get in touch with their children or contact the police. Of course, in addition to the elderly group, you yourself or your friends and relatives received this kind of help video, the best choice is to call the police. The police can identify whether it is true or not.
The lack of ethical constraints on technology research, which ignores the huge risks behind it, is mostly for the short-term interests of a few people, not for the so-called “development of society”. People with bad intentions will use AI face-swapping technology to do things that harm society, but we can’t completely deny the value of this technology. We should think about how to use it in the right way and play its positive role.