On January 12, the People's Think Tank and the Megvii AI Governance Research Institute jointly released the "Top Ten Global Artificial Intelligence Governance Events in 2021" (hereinafter referred to as the "Top Ten Events of AI Governance"), covering major legislative practices, ethical norms formulation and cutting-edge scenarios in the field of artificial intelligence at home and abroad in the past year.
Whether it is the human brain chip is expected to land this year, automatic driving or will be legalized in the United Kingdom, or the final judgment of the domestic "face recognition first case", and the united Nations passing the first global agreement on AI ethics, all of them show people's ambivalence of both desire and vigilance in AI technology. The urgency of AI governance speaks for itself.
Some experts told Nandu reporters that the active legislation of AI at home and abroad means that AI applications have entered the popularization stage, and it is urgent to formulate relevant systems and norms. Some experts remind that AI is not only an important means to create a better life, but also an object that needs to be properly governed.
1
Ai applications have entered the popularization stage, and legislation has been enacted at home and abroad
Nandu reporter learned that the "Ten Major Events of AI Governance" was jointly established by the People's Think Tank and the Megvii AI Governance Research Institute, invited the Chinese Intelligent Industry Development Alliance (AIIA) as the project support unit, and organized authoritative experts and scholars in related fields and practitioners in the field of artificial intelligence industry to jointly select and select.
The application of AI is inseparable from big data, and ensuring data security can better carry out AI governance. Therefore, the first of the ten major events is the successive release and implementation of domestic data security-related legislation.
On August 1, 2021, the first judicial interpretation on face recognition in China, the Provisions of the Supreme People's Court on Several Issues Concerning the Application of Law in the Trial of Civil Cases Involving the Use of Face Recognition Technology to Handle Personal Information, came into effect. A month later, the Data Security Law of the People's Republic of China came into effect. On November 1 of the same year, the Personal Information Protection Law of the People's Republic of China was officially implemented, and in the same month, the Regulations on the Administration of Network Data Security (Draft for Solicitation of Comments) were released.
Wei Kai, deputy director of the Institute of Cloud Computing and Big Data of the China Academy of Information and Communications Technology, pointed out that the Data Security Law, the Personal Information Protection Law and the Cybersecurity Law have jointly formed a "three carriages" in the field of data governance law, marking the initial completion of China's data security legal framework.
In addition, 2021 is also known as the "first year of China's algorithmic governance". In August last year, the Cyberspace Administration of China (CAC) solicited public opinions on the Provisions on the Administration of Recommendation of Internet Information Service Algorithms (Draft for Solicitation of Comments), proposing that users should be provided with options that do not target their personal characteristics, or provide users with convenient ways to shut down. Subsequently, nine ministries and commissions, including the Cyberspace Administration of China, issued the Guiding Opinions on Strengthening the Comprehensive Governance of Internet Information Service Algorithms, announcing that China will establish a comprehensive algorithm management system within three years. At the beginning of 2022, the Provisions on the Administration of Recommendation of Internet Information Service Algorithms were promulgated and came into effect in March.
"The comprehensive management of algorithms in 2021 and related legislation have enabled China to take the lead in establishing the world's most complete Internet governance system including data security, personal information protection and algorithm security." Zhang Linghan, an associate professor at the School of Law at the University of Science and Technology Beijing, said.
In terms of judicial practice, in April 2021, the "first case of face recognition" ushered in the final judgment. The plaintiff, Guo Bing, argued that The Hangzhou Wild Animal Park forced annual pass users to swipe their faces to enter the park infringed on the rights and interests of consumers, and sued him in court. In the end, Hangzhou Wildlife World was ordered to delete the information on facial features, including photos, and fingerprint identification information submitted by the plaintiff when applying for a fingerprint annual card.
Liu Wei, director of the Human-Computer Interaction and Cognitive Engineering Laboratory of Beijing University of Posts and Telecommunications, analyzed that the reason why the "first case of face recognition" has attracted much attention is because it reflects the increasing attention of the public to face recognition information; it also reflects the protection of personal information at the judicial level, which has paid full attention to the three principles of "legal, legitimate and necessary" for enterprises to collect and use personal information; it has also sounded the alarm bell for enterprises, reminding enterprises to pay more attention to legal compliance in the collection and use of personal information.
Stricter regulations and accelerated legislation have become a global trend in AI governance. On April 21, 2021, the European Union released for the first time a draft regulation on AI technology, which is considered to be the most stringent comprehensive regulatory regulation on AI technology in the world. On 6 October 2021, the European Union passed a resolution banning the use of facial recognition technology by police in public places and imposing strict restrictions on its use of AI for predictive policing activities.
"Ai-power governance has become a global consensus, and has entered the stage of establishing rules and regulations and implementing them from the conceptual level." Liang Zheng, professor at tsinghua university's School of Public Policy and deputy dean of the Institute of International Governance of Artificial Intelligence, said that active legislation on AI at home and abroad means that AI applications have entered the popularization stage, and it is urgent to formulate relevant institutional norms, in addition to more detailed rules and guidelines for subdivision scenarios to promote the implementation of laws and regulations.
2
Flexible ethical norms are the other side of AI governance
In AI governance, rigid legal constraints and flexible ethical norms are often parallel. Since the birth of AI technology, the complex ethical issues it contains have been highly concerned and controversial from all walks of life. From the perspective of the "Top Ten Events of AI Governance", in addition to law, ethics has also become the keyword of AI governance in 2021.
On September 25, 2021, the National New Generation Artificial Intelligence Governance Professional Committee issued the "New Generation artificial intelligence ethics code", which put forward six basic ethical requirements such as improving human well-being, promoting fairness and justice, protecting privacy and security, ensuring controllable and credible, strengthening responsibility, and improving ethical literacy, aiming to integrate ethics and morality into the whole life cycle of artificial intelligence, and provide ethical guidance for natural persons, legal persons and other relevant institutions engaged in artificial intelligence-related activities.
Looking at the world, on June 28, 2021, the World Health Organization officially released the "Guidelines for the Ethics and Governance of Artificial Intelligence in the Field of Health". On 25 November 2021, UNESCO held a press conference to announce the official adoption of the Ai Ethics Recommendation, the first global agreement on the ethics of AI.
The "Ai-FiEn Ethics Recommendation" proposes that the development and application of AI should first reflect four major values, namely, respecting, protecting and enhancing human rights and human dignity, promoting the development of the environment and ecosystems, ensuring diversity and inclusiveness, and building a peaceful, just and interdependent human society.
Compared with laws and regulations, ethical norms reflect a general direction and universal guidance. Enterprises are the protagonists of the implementation of ethical norms - in recent years, the establishment of ethics-related departments has become an important manifestation of corporate self-discipline.
For example, in 2017, Microsoft established an internal Artificial Intelligence Ethics Committee (AETHER); Megvii, which initiated this selection, established an "Artificial Intelligence Ethics Committee" in 2019 and established a "Megvii AI Governance Research Institute" in early 2020. In recent years, calls to strengthen ai ethics have also increasingly come from companies.
At the two sessions of the National People's Congress in 2019, Baidu founder Robin Li proposed to accelerate ai ethics research and encourage enterprises to implement AI ethical principles in product design and business operations. At the 2020 World Artificial Intelligence Conference, Megvii co-founder and CEO Yin Qi proposed three principles that enterprises should uphold when practicing AI governance: not absent - deeply involved in it; not opposing - the development of technology, the application of business, and the formulation of rules should complement each other; actions speak louder than words - every step of AI governance should be reflected in daily work. At the 2021 World Artificial Intelligence Conference, Tao Dacheng, president of the Jingdong Exploration Research Institute, said that the choice problems faced by AI in the landing application are mainly manifested in the crisis of public trust in AI technology and ethical anxiety.
According to Yang Danhui, director of the Resources and Environment Research Office of the Institute of Industrial Economics of the Chinese Academy of Social Sciences, AI technology is still in the early stage of industrial application, and many deep-seated ethical problems and their impact have not yet been fully revealed. Therefore, relevant departments should closely follow the forefront of technology, widely absorb the opinions and suggestions of experts and scholars in different disciplines and fields, as well as enterprises and consumers, and scientifically and dynamically adjust ethical norms.
In addition, AI governance should become a broad consensus on "multi-party co-governance". Liang Zheng analyzed the "Ten Major Events of AI Governance" and said that in the governance of AI, in addition to the top-down establishment of rules and regulations, bottom-up participation based on application subjects and user choices has also become an important feature of AI governance.
"For example, this time we see that in the top ten events of AI governance, some companies provide some automated decision-making and information selection rights in their product applications, and return them to individual consumers." Of course, this aspect will also bring some controversy, but it reflects our concept of 'multi-party participation and collaborative governance' in AI governance. He said.
3
When using AI, you need to consider its limitations at the same time
In the past year, the corresponding regulatory implementation has been the imagination of the future brought to the world by technology.
In the "Top Ten Events of AI Governance", events such as the legalization of self-driving cars in the UK in the short term, human brain chips or will be put into use in 2022 have made people see the technological boom. However, experts generally agree that these cutting-edge technologies need to be viewed more carefully.
Taking the legalization of autonomous driving as an example, Ning Yang, a professor at North China Electric Power University, said: "This aspect marks the increasing maturity of autonomous driving technology, which may make future travel more green, convenient and reliable, and will promote the revision of related regulations such as road traffic and commercial insurance." But on the other hand, it will also bring people's concerns about the safety of automatic driving and the loss of driving experience. This is undoubtedly a new technological revolution. ”
In fact, whether technological progress will bring more well-being or more harm to mankind is also a matter of general concern and concern.
For the more science fiction issue of human brain chips, Tang Lei, a researcher at the Institute of Political Science of the Chinese Academy of Social Sciences, believes that this is a scene that has been repeatedly conceived in science fiction and movies, and it is a key step to break through the limits of people. "The field of health care is the initial, most direct and most important application area of brain-computer interface, which will bring great benefits to human beings. However, if the brain-computer interface is used in the military field to replace and enhance the direction, it may bring great disasters to mankind. ”
The "Top Ten Events of AI Governance" also pays attention to the further extension of ai landing scenarios. In the past year, Zhu Tingshao, a researcher at the Institute of Psychology of the Chinese Academy of Sciences, developed a suicidal ideation recognition model through AI algorithms and organized a team of volunteers to intervene and rescue. The first AI sign language anchor of CCTV News and Baidu Intelligent Cloud was officially unveiled; iFLYTEK and iQiyi have also launched simulated AI anchors that are good at sign language expression. On Earth Day, Alibaba Damo Academy, Nanjing University of Information Science and Technology, National Climate Center, etc. jointly held the industry's first AI climate prediction competition, predicting the climate for the next two years for the first time.
"The continuous development of AI technology inspires different industries, providing new ideas for many previously difficult problems to solve, and the scientific and rational use of AI capabilities can benefit human society." Li Jinlong, director of the artificial intelligence laboratory of China Merchants Bank, said. However, he also reminded that when using AI itself, it is also necessary to consider the limitations of AI itself - "The bottom layer of today's AI technology is still statistical thinking, and we must pay attention to the problem of AI system bias in actual use."
Wang Aihua, deputy chief engineer of the China Academy of Information and Communications Technology and deputy secretary-general of the Chinese Intelligent Industry Development Alliance, believes that the ten major events show that AI governance is "becoming stricter, more realistic, and more refined", and the legislative process at home and abroad shows that AI governance is ushering in a substantial regulatory landing, and the focus of governance is refined in typical scenarios, such as automatic driving, smart medical care, etc.
Since the first selection of the world's top ten AI governance events in 2019, Yin Qi has participated in the selection work every year. In his view, a significant trend this year is that the laws and regulations related to AI governance, algorithm governance, and data governance are more perfect, and more refined standards and guidance are given to the subdivision scenarios of AI technology and application.
As a key area of global science and technology competition, AI not only needs to work technological innovation and application, but also needs to exert efforts in governance. Artificial intelligence is not only an important means for us to create a better life, but also an object that needs to be properly governed. He said.
Written by: Nandu reporter Li Yaning