The invention of emerging technologies such as artificial intelligence, big data, and 5G has added a lot of impetus to the development of society, but it is also a serious threat to the security of the entire country. Deep learning under artificial intelligence seems to be a normal branch of science, but it has quietly created an opportunity for deepfake technology. Based on the background and characteristics of the rise of deepfakes, this paper analyzes the instrumentalism it plays in various serious criminal offenses, revealing that it has brought many dangers and social challenges to the field of national security. In response to deepfakes, it is necessary to strengthen the research and development of early warning and detection technologies, continuously strengthen legislation and supervision, and emphasize the joint construction of a national security community in promoting international security cooperation and governance.
With the research and development of Web 4.0 technology, breakthroughs have been made in deep learning and algorithm innovation in the network field, and artificial intelligence technology has also gained more new inspiration in the third wave. Artificial intelligence technology has ushered in a new round of implementation and reform, but with the gradual arrival of risks, deepfakes are gradually approaching people's field of vision. In January 2019, the Carnegie Endowment for International Peace Research published an article titled "How Countries Can Respond to the Deepfake Chaos", which clearly pointed out that the damage caused by deepfake technology to governments and enterprises is far-reaching, and the impact is also negative. The fact that the U.S. authorities held hearings on the issue of deepfake technology is enough to show that deepfake technology has an important and far-reaching impact on social development and change. This is of indispensable importance for deepfake technology, but due to technical problems, it has not been able to attract enough attention from the top level of the United States. Through the background and characteristics of the rise of deepfake technology, we can deeply analyze the new challenges it brings to the country's politics, economy, ecology, culture and other aspects, and find reasonable countermeasures from it while further analyzing the risks and coping with the challenges.
1. The background and characteristics of the rise of deepfakes
(1) The background of the rise of deepfakes
The concept of deepfakes first appeared in late 2017, when users named "deepfakes" gained attention by posting an indecent video online that was spliced together. Deep fake is a combination of deep learning and fake, using artificial intelligence technology to deepen the application of deep learning portrait synthesis technology. With the vigorous development of network technology, deepfake technology has become very mature in the deep synthesis technology of video, voice, text and micro-expressions, and has reached the point where it is fake. Behind the deepfake technology is the further research of deep learning, with generative adversarial networks (GANs) and convolutional neural networks (CNNs) as the main contents. Moreover, the more profound the relevant data of the learning algorithm, the more real the data will be in the actual application, resulting in the higher the degree of realism and simulation of the product content synthesized and forged.
(2) The technical principles of deepfakes
1. Video forgery: AI face swapping
As the core typical representative of deepfake technology, video forgery is also known as AI face swap, that is, AI face swap. It uses technology to generate the corresponding adversarial network or convolutional neural network, and on this basis, the human face to be imitated and faked is "grafted" and "transplanted". But in order to form a complete fake video, you need to connect each spliced picture into a complete fake video after technical processing, and finally what you see is the stitched video after technical processing. And such a technology is not difficult for continuous learners in the early stage, so more and more AI face-swapping videos are circulating on the Internet.
2. Audio Forgery: Speech Synthesis
Another typical type of deepfake is audio fakes, which are grafted using "voice cloning," which originally appeared as virtual items. Based on this, speaker adaption and speaker encoding are used as typical representatives for speech cloning, and the forged speech is spliced. Moreover, each person's pitch and tone are different, and speaker coding is to use technology to fine-tune the speaker's voice, and constantly adjust the type of voice to be imitated in the fine-tuning, and then imitate and fake a small number of speakers' voices, and find the characteristics of the sound in continuous imitation.
3. Other forgery techniques: splicing with each other
Although splicing and grafting are common in video and audio tampering, this does not include deepfakes for other audio and video types. At present, the main technical support of deepfakes lies in the difficulty of distinguishing between true and fake in natural language processing (NLP), which is mainly due to the limitations of network technology and the limitation of identification technology, which leads to criminals using natural language processing technology to generate fake texts equivalent to video and audio. The micro-expression forgery uses image synthesis technology to instantly capture the face of the forged object, and after technical analysis, the facial micro-expressions in the face are further processed, and the face is further judged and processed in nearly 40 kinds of facial micro-expressions, so as to realize the deep-fake of the facial micro-expressions of the human face.
(3) Characteristics of deepfakes
1. It is mainly based on algorithm technology, and the cost and threshold are extremely low
In fact, counterfeiting technology has appeared before, but there is an essential difference between traditional counterfeiting and artificial intelligence counterfeiting. Some puzzle technology software in daily life can be easily grafted without having enough deep learning algorithm skills, and it is for this reason that counterfeiters can further realize what they want to fake in deepfakes. The technical threshold for such software is extremely low, and relevant real-world cases can be searched on the Internet, and deepfakes can be easily achieved without the cost of counterfeiting. In addition, with the further proliferation of computing technology and the programmatic use of forgery technology, the learning of deepfake technology no longer exists. Some computer enthusiasts can see some fake video learning materials on code-hosted websites. Therefore, at that time, it appeared for a long time, and some of the APP of face swapping software was available for download and use, and under the effect of photos, it was easy to convert and replace. In other words, whoever masters the production and use of these software and materials will have a breakthrough at the core of the entire deepfake technology, and this process is a highly technical process.
2. The effect is realistic and diverse, and it is difficult to distinguish
Since deepfake technology covers multiple types of fakes including video, audio, text, micro-expressions and many other common content, deepfake technology in a wide variety of situations may pose a greater threat to the overall security of the entire country, and this threat will further spread in such an environment. Deep learning can extend and expand deepfakes to a large extent, and self-learning can be achieved in repeated loops. Specifically, under the study of clear goals, fake electronic products such as fake video and audio will be forged in terms of content, making it impossible to distinguish the authenticity from the fake. At present, deepfakes are mainly concentrated among celebrities, politicians and other public figures, and it is precisely because the public can obtain relevant information at will that it has become the raw material of deepfakes. For the average person, some facial features can be observed with the naked eye at the beginning of deepfake products, but with the continuous advancement of this type of technology, it is no longer possible to observe with the naked eye when the deepfake technology is more mature in 2017. For a long time to come, a variety of morphological features such as voice, micro-expressions and text have become a trend in counterfeiting technology, and it will be more difficult for the masses to self-identify.
3. With the network as the carrier, the propagation speed is fast
In terms of the production of deepfake technology, it is a considerable number of people's curiosity to be satisfied, and many netizens are constantly satisfied with showing off themselves, so that people all over the world know their faces. The development of the Internet as a whole has gradually become a whole in the changes of the global network, and it is precisely under such an environmental background. Deepfake products use their powerful sphere of influence to gradually become a carrier of cyberspace development in the new era in the scale of change. At the same time, the wider the reach of counterfeits, the greater the number of clicks they attract, and the rapid spread of counterfeit products can cause catastrophic damage to the overall operation of national security. Cyber changes are fast-moving, and without causing serious damage, they are already causing widespread spread. A large number of AI face-swapping incidents and indecent video splicing and other prominent problems have brought some negative negative impacts on foreign websites.
2. Challenges to national security brought by deepfakes
(1) Deepfakes impact economic security
1. Deepfakes become accomplices in ruining the corporate image
As an important part of the socialist market economy, once deepfakes occur within the enterprise, it will cause serious damage to the development of the enterprise itself, which is not conducive to the good image of the enterprise in the global economic construction. In the case of business competition between companies, once a deepfake attempt is made to smear competitors, it may bring a greater crisis of trust to the other party. The content and introduction of the company's executives and relevant management personnel are collected and then synthesized into fake videos and audio to discredit the person in charge of the company, which continuously undermines the company's business reputation and causes damage to the company's reputation. In the future, if deepfake technology deliberately slanders the other company's company by producing fake videos and fake audio, it will cause serious damage to the corporate image and is not conducive to the development of the socialist market economy.
2. Deepfakes pose new risks to the stability of financial markets
To a large extent, deepfakes will cause turmoil in the entire financial market and provide the soil for the survival of more unstable development factors in the future. For example, the judgment made by relevant economists, experts and professors at home and abroad on the direction of economic development in the coming period is usually distorted and misled by deepfakes with ulterior motives by producing fake videos and fake audio, thus causing panic and causing serious damage to the reputation of enterprises or businesses, and the culprits will sit back and enjoy the results. It can be seen that due to the serious impact of the so-called authoritative judgment being falsified, it has directly led to serious damage to its own economic interests.
(2) Deepfakes endanger social security
1. Cause a crisis of social trust
As far as the crimes committed by deepfake technology are concerned, the fundamental social crisis lies in the crisis of trust, and many social problems between people are caused by distrust between each other, which over time leads to a negative impact on interpersonal interactions between people. However, in interpersonal communication, the traditional view that seeing is believing and hearing is still an important basis for judging interpersonal relationships, and if deepfakes are inserted into it, it will inevitably seriously affect the interpersonal communication of the whole society. The development and application of face swapping software and voice changing software, due to the extremely low production cost, directly leads to the embarrassment of interpersonal communication in society. Once fake video and fake audio flow into society, due to the extremely fast transmission speed, it directly leads to the spread and spread of social trust crisis, and the so-called truth falls into an abyss that will never recover, making society enter the "post-truth era". The construction of socialist democracy and the rule of law has been seriously trampled on, and loans, credit, and other types of business may not be able to be handled, and social development after losing trust will be in a state of stagnation, which is the crisis of trust under deepfakes.
2. Challenge existing legal boundaries
Deepfake technology, which automatically generates high-fidelity fake videos and audios, can cause a lot of trouble for society as a whole if it is not used unchecked. The law attaches great importance to the in-depth integration between theory and practice, everything is based on facts, the law is the criterion, and under the support and guidance of the facts and evidence of the case, it brings vitality to the future development and construction of the socialist rule of law. However, under the rapid spread and development of deepfake technology, social development will also stagnate without the support and guidance of technology, but as far as the construction of the socialist rule of law system is concerned, the emergence of deepfake technology will also bring related legal risks: one is framing others. One is to provide false evidence to conduct false litigation. There is also the category of inability to find the initiator in the constant forwarding, dissemination and diffusion. With the rapid development of technology and the continuous integration with the world in keeping pace with the times, there will be mistakes in the identification and judgment of deepfake technology, and the truth of the case and social fairness and justice will be buried by deepfake technology. In view of the flexible mastery and application of deep learning and deepfakes, the law should be further strengthened and strengthened in order to crack down on deepfake technology.
3. It is easy to breed terrorism
National security is still facing a serious threat, deepfake technology has quietly become the secret weapon of terrorists, but the goal of its implementation and attack is to cause serious social panic to the whole country, to achieve its own illegal purposes, through deepfake technology to achieve terrorist goals. Once more terrorists have mastered the forgery technology, it will be more convenient for them to carry out terrorist acts, which is actually a serious trampling on the beautiful development of socialism and economic construction, and is not conducive to the state's severe crackdown on violent and terrorist criminal activities. If terrorists and deepfake technology "join forces", then it will have a negative impact on the overall effect of national governance and social governance, and there will be more incidents that endanger national security between "cyber terrorism" and "deepfake terrorism", which requires the prevention and containment of the spread of terrorism at the same time, highlight the control and suppression of deepfake technology, otherwise it will inevitably pose a serious threat to national security.
(3) Deepfakes affecting national peace and security
1. Threatening the lives of the public
The large-scale dissemination of fake videos and audios based on deepfake technology will pose a serious threat to the lives and health of the general public. Therefore, with the continuous development of deepfake technology, criminals use these fake videos and fake audio to commit illegal crimes, which will pose a serious threat to the lives and health of the general public.
2. Causing property losses to the masses
Deepfake technology can easily produce fake videos and audios, and provide false evidence and proof for other individuals and organizations, which can easily breed illegal and criminal behaviors and lead to serious threats to the property security of citizens. Speech as an essential feature of human pronunciation, although it is highly recognizable in terms of biological characteristics, it is difficult for ordinary citizens to identify due to the continuous improvement of forgery technology. Some scam calls will take the opportunity to defraud the victim and are not fundamentally different from the person being imitated. Among some of the cases seized by the public security organs, criminal cases of criminals using fake videos and audio to engage in telecommunications network fraud have emerged one after another. It can be seen that with the continuous improvement of deepfake technology, it has posed a serious threat and challenge to the safety of people's lives and property. What's more, in kidnapping and extortion cases, this technology is constantly used to defraud the people's vital property, so that the victim's family is also implicated, because it is impossible to distinguish between the real and the fake, and they are worried that their own family members will be implicated, so it has led to the frequent occurrence of various kinds of cases, and has also brought resistance to the severe crackdown on crime and crime prevention.
3. Detrimental to the personal reputation of the masses
Deepfakes are also prone to have a serious negative impact on an individual's reputation and spirit. Since deepfakes were transplanted and grafted onto indecent videos, it has become a new type of weapon to discredit others to some extent. Because the faked situations and scenes are very realistic, it once made it difficult for relatives and friends who knew them well to believe their eyes, and in some extreme cases, it once caused the victim to be mentally anxious and even commit suicide. Mental torture is the greatest pain for a person, and the sophistication of deepfake technology can cause more serious reputational damage to the entire country, society and people. If you use splicing and grafting to forge other people's avatars for slander or insult, it will lead to the breakdown of your own family and the reduction of social trust. In addition, it will also pose a threat to national security to varying degrees. At present, fake videos and fake audios in the nature of online spoofs will cause trampling on the personal reputation of others. Artificial intelligence face-swapping technology is a technical refinement between deep learning and deepfake, and such technology is difficult to decipher. Although the content of the falsified characters is fake, it has a negative impact on the reputation and social status of individual citizens. Everyone lives in society and plays a different role in society, and reputation often represents a social trust and status, so it is of great significance to individual citizens.
III. Some Thoughts on Preventing and Responding to Deepfakes Endangering National Security
(1) Strengthen exchanges and cooperation between various departments
The risk brought by deepfakes to national security is not inferior to traditional state slander, and the seriousness and urgency of deepfakes to national security are real problems faced by all countries. First, it is necessary to establish a global outlook on national security. At present, scientific and technological inventions are changing with each passing day, which stimulates the development concept of various countries to a certain extent. Building a community with a shared future for mankind requires mutual cooperation among various countries, so in the development of advanced technologies such as artificial intelligence, comprehensive exchanges and cooperation should be carried out around the essence of national and social development. Artificial intelligence is indispensable for the future development of a country, and it also provides certain convenience for the development of deepfake technology, which has caused serious threats and challenges to the entire national security. The second is to ask the government to give the greatest help and support in the fight against deepfakes. Bringing together police, courts, and scientific research and technology personnel to increase research and development efforts in early warning and detection technology tools, and continuously improve the continuous crackdown on deepfake technology. It is of epoch-making significance to prevent and resolve deepfake technology by improving the intensity of the crackdown and adopting the method of "countering violence with violence" and looking for differences in the clues. Third, the government should encourage scientific and technological innovation. At the same time of technology research and development, we will integrate technological innovation, give full play to the role of universities and research institutes in technology research and development, and continuously improve the crackdown and response to non-traditional national security factors.
(2) Jointly advance criminal legislation and administrative supervision
First, in view of the legal liability of deepfake products among the production entities, it is necessary to increase the promotion of criminal legislation from the source. Deepfake technology itself is meant to serve the state, society, and individuals, and in the face of criminal cases involving the use of deepfake technology, it is necessary to increase the promotion of criminal legislation. Strengthen and restrain the moral and discipline work styles of government officials themselves, and do not leave opportunities for people with ulterior motives to take advantage of. The second is the supervision and management of the responsibility of online media, and in the process of dissemination, it is necessary to report real news content and convey positive social energy. The rapid fermentation of online media will have a fundamental impact on a country's institutional concept, and the speed of communication will directly lead to the situation getting out of control. The cyber world, like social development, is an important factor influencing the progress of a country's economy and political civilization. At the level of administrative regulation, deepfake technology should be restricted first, and even if it is entertainment in nature, it should be carried out under the premise of clear notification, and explanations should be given for those that may cause political turmoil. In order to highlight the responsibility of the responsible entity, it is necessary to adopt criminal legislation to promote the industry supervision of deepfake technology after going beyond the administrative regulation. Third, it is necessary to establish a corresponding response mechanism, and once a person or department suspected of deepfake is discovered, it is necessary to avoid the expansion of the situation as much as possible. To this end, government departments should continue to coincide with their own national interests in international exchanges, and actively carry out relevant cooperation and exchanges under the condition of ensuring the smooth flow of technology and information.
(3) Raise citizens' awareness of self-prevention
First, it is necessary to increase the education and active guidance of its own citizens, continuously improve citizens' awareness of self-prevention, and make a centralized display of common types of deepfakes to prevent the masses from being deceived. Party committees and government departments at all levels shall continuously raise their awareness of prevention in the course of television propaganda and thematic education, and should not make judgments in the traditional form of "seeing is believing" or "hearing is believing" in the past, but should improve their own precautions in the course of self-judgment. In some cases of telecom network fraud, deepfakes often appear. Second, it is necessary to provide citizens with more prevention skills. With the continuous deepening of deepfake technology, substantive judgments should be made on the authenticity of some case information. At this time, it is necessary to cooperate continuously between the government and its functional departments. We should intensify scientific and technological research and development and tackle tough problems, provide more technical support for the broad masses of the people, and avoid being deceived and suffering more economic losses. Third, it is necessary to correctly guide online media to make more positive reports. With the proliferation of deepfake technology in the era of self-media, people can watch and spread through some social networks or self-media. Based on this, deepfake technology uses fake videos and fake audios to spread and spread at the same time, which may cause deep transmission and spread. Therefore, it is necessary to guide and educate the broad masses of the people to unconsciously identify the true from the false, and to prevent them from becoming "accomplices" in the process of reading.
(4) Highlight China's participation in China's smart China plan
First, deepfakes should be incorporated into the Global Governance Cooperation Framework process. The further development of deepfake technology will cause serious damage to the entire national security, and the proliferation of deepfake products has become a bottleneck that is difficult to effectively govern. Deepfake technology is a replica of artificial intelligence technology, and some countries outside the region have clearly pointed out that there is a risk of abuse of artificial intelligence technology, so deepfake technology has become a model among them. Therefore, the state should continue to deepen the impact of combating and preventing the spread of deepfake technology on national social development, and prevent the further spread and spread of deepfake technology. Second, it is necessary to strengthen international cooperation in combating deepfake technology, especially transnational crimes and transnational terrorist crimes. In the eyes of Western countries, criminals committing illegal and criminal acts through the use of deepfake technology is a major trend of new crimes. Moreover, this trend is gradually developing into a severe crackdown on crimes that endanger national security, coupled with the use of deepfake technology by criminals to engage in acts that endanger national security. In order to effectively deal with this risk, a reciprocal mechanism should be adopted, and China's wisdom should be used to highlight China's plan and governance for deepfakes. Third, in view of the legal responsibility and social responsibility between the enterprises involved, it is necessary to strengthen the strict implementation and control of the main responsibility. At present, the development and development of social software has made more and more beneficiaries, but we must be vigilant and beware of the negative negative effects arising therefrom. It is necessary to strengthen cooperation among countries to prevent the negative impact of more false news on national development.
epilogue
The overall national security concept is of great significance in preventing and responding to the illegal crimes committed by deepfake technology. The overall national security concept points out the direction and path for further eradicating deepfake crimes, and avoids the occurrence of more and greater deficiencies. It is necessary to raise the awareness of identifying and preventing deepfake incidents, which will provide directional guidance for further development and enhancement of the country's efforts to combat deepfake crimes, which is of epoch-making significance for safeguarding national security and territorial sovereignty integrity.