Looking back five years ago, the topic of security application of artificial intelligence technology has entered the public's vision.

At the 2018 Artificial Intelligence Conference, Li Yanhong proposed to "follow AI ethics". He believes that real AI companies are not only closely integrated with AI at the technical level, but also culture must be "AIized". He said that the four major principles of AI are safety and controllability, promoting equal access to technical capabilities by human beings, teaching people to learn to make people grow, and bringing more freedom and possibilities to mankind. Among them, "safety and controllability" is the highest principle of AI. .

It was also at the Artificial Intelligence Conference that the "Shanghai Initiative for the Security Development of Artificial Intelligence" was released, proposing eight aspects, including future-oriented, clear responsibilities, transparent supervision, privacy protection, fair algorithms, peaceful use, and international cooperation.

In May 2019, the Shanghai National New Generation Artificial Intelligence Innovation Development Pilot Zone was unveiled, clarifying the relevant tasks of establishing and improving policies, regulations, ethical norms and governance systems. In June, the National New Generation Artificial Intelligence Governance Professional Committee released the "Next Generation Artificial Intelligence Governance Principles", highlighting the theme of responsible and open collaboration. At the 2019 World Artificial Intelligence Conference Governance Theme Forum, the "2019 Shanghai Declaration on the Governance of Artificial Intelligence by Young Chinese Scientists" was released, emphasizing the four major responsibilities that need to be followed in the development of artificial intelligence, including ethical responsibility, security responsibility, legal responsibility and social responsibility, etc. .

In 2020, the new generation of artificial intelligence technology will move from "perceptual intelligence" to "cognitive intelligence", allowing machines to achieve understanding, reasoning, decision-making, and interpretable cognitive intelligence is becoming a hot topic and trend in AI research. Against the backdrop of further upgrading of machine intelligence, how to develop "responsible artificial intelligence" has become an important topic. At the World Artificial Intelligence Conference that year, multiple forums discussed the issue of AI's rule of law, security and ethics.

In 2021, it has become a general trend to strengthen the supervision of artificial intelligence governance and accelerate legislation.

On the one hand, my country's legislation on network security, data security, and personal privacy protection has been gradually improved. In August 2021, the first judicial interpretation of facial recognition in China, "Regulations of the Supreme People's Court on Several Issues Concerning the Application of Laws in the Handling of Civil Cases Related to the Use of Face Recognition Technology to Handle Personal Information" was officially implemented; in September, "Data of the People's Republic of China The Security Law was implemented; on November 1 of the same year, the "Personal Information Protection Law of the People's Republic of China" was officially implemented; in the same month, the "Regulations on Network Data Security Management (Draft for Soliciting Comments)" was released. All parts of the country have also accelerated the process of data legislation and construction of data trading venues.

On the other hand, my country has further promoted algorithm governance and issued the "Guiding Opinions on Strengthening the Comprehensive Governance of Internet Information Service Algorithms", and the "Regulations on Recommended Management of Internet Information Service Algorithms" was also officially implemented in March 2022.

Also in this year, scientific and technological ethics issues have ushered in centralized governance.

In November 2021, the 41st session of the UNESCO General Assembly reviewed and passed the "Recommendation on the Ethics of Artificial Intelligence", which is the world's first global agreement on the ethics of artificial intelligence, calling on governments to accelerate capacity building and promote scientific and technological development. We call for full respect for human rights and human dignity, environmental protection and sustainability, and commitment to creating a fair and inclusive society, we call for the formulation of paths for the development of artificial intelligence, and pay attention to the entire life cycle of artificial intelligence, etc.

"The main risk brought by artificial intelligence at present is not necessarily the immaturity of technology, but the impact of the rapid development of AI technology on ethical and social culture has exceeded our understanding." At the 2022 World Artificial Intelligence Governance Forum, Hong Zhiwei, the UNESCO representative office in China, said that it is necessary to reduce the social and ethical risks brought about by the use of AI technology, such as aggravating inequality, privacy violations, and causing prejudice. The embedded governance of values ​​urgently requires international cooperation to form multilateral and flexible AI governance policies and achieve more predictive AI governance.

At the same time, my country's scientific ethics governance process is also accelerating. In March this year, the "Opinions on Strengthening the Governance of Science and Technology Ethics" was officially issued, further emphasizing the need to improve the science and technology ethics system and strengthen the main responsibility of innovation entities in science and technology ethics management.

"The Opinions on Strengthening the Governance of Science and Technology Ethics" has promoted my country's science and technology ethics system to enter a new stage." At the Asian Science and Technology Ethics Governance and Sustainability Forum, President of the Shanghai Modern Service Industry Federation and former Deputy Director of the Standing Committee of the Shanghai Municipal People's Congress at the Asian Science and Technology Ethics Governance and Sustainability Forum, the President of the Shanghai Modern Service Industry Federation and former Deputy Director of the Standing Committee of the Shanghai Municipal People's Congress Zheng Huiqiang pointed out that we must abandon the model of development first and then governance, but should adhere to the model of development in normative and standardized in development, and explore new governance models under the issue of sustainable development between government, industry, academia and research.

In order to further build the "Shanghai Highland" for the development of the artificial intelligence industry, Shanghai has also fully considered the AI ​​ethical system in terms of artificial intelligence legislation. On August 30 this year, before the opening of the 2022 World Artificial Intelligence Conference, the "Shanghai Regulations on Promoting the Development of the Artificial Intelligence Industry (Draft)" was officially solicited for comments, which mentioned the establishment of an artificial intelligence ethics committee, advocated innovative hierarchical governance, and developed The ethics of the company are standardized, requiring enterprises to reasonably apply artificial intelligence technology, strengthen workers' protection and care for special groups.

At the theme forum of "2022 World Artificial Intelligence Conference Rule of Law Forum - Digital Rule of Law: Supply and Applicability", relevant Shanghai departments leaders introduced three key factors regarding Shanghai's artificial intelligence legislation, including leading industrial innovation and deepening reform and development with good laws and good governance, using good laws and good governance to lead industrial innovation and deepen reform and development , solidify relevant industrial practical experience into legal system provisions, and promote the unity of artificial intelligence technology innovation and risk prevention. Especially in terms of legislation to ensure industrial security and strengthen standardized governance, Shanghai regulations are foreseeable and are committed to balancing the relationship between industrial development and safety norms. While regulations are made, they are also used to prepare for new tracks and new industries within a controllable range. Leave room for development.

A discussion on the "embedded" technology ethical governance model

"With innovative breakthroughs and accelerated application of artificial intelligence technology, three new features have emerged in the governance of artificial intelligence. First, artificial intelligence is accelerated application, resulting in more complex and diverse governance problems, and the dimension and breadth of governance are constantly expanding. Second, A more systematic artificial intelligence governance model is being formed, and the methods and means of governance are becoming increasingly rich. Third, international exchanges and cooperation in artificial intelligence governance are more active, and we actively promote the formation of international consensus and international mechanisms for artificial intelligence governance." Strategic Planning of the Ministry of Science and Technology Deputy Director of the Department Xing Huaibin pointed out in his speech at the 2022 World Artificial Intelligence Conference Governance Forum.

Under the guidance of the government and industry associations, the AI ​​ethical governance of enterprises has been raised to a new level, and establishing an "embedded" technology ethical governance system may be a major direction.

The so-called "embedded" science and technology ethics governance means that in the early stages of scientific and technological innovation, science and technology ethics should be intervened to effectively intervene and manage the development and risks of emerging technologies.

"With the explosive growth of the AI ​​industry, AI leaders have begun to realize that before technology is put into use, ethics must be taken into consideration and AI governance principles need to be put into the entire deployment process." Aopeng Appen Global Tian Xiaopeng, senior vice president and general manager of Greater China and North Asia, pointed out that if responsible AI is to be deployed, responsible data must be available. Aopeng Appen has provided AI training data services worldwide for more than 26 years. In an exclusive interview with a reporter from 21st Century Business Herald, Tian Xiaopeng mentioned that machines are a supplement to human labor, but machine training still requires a lot of manpower, especially expert abilities. , How to correctly train computers with high-quality data and promote computer learning and development is a major difficulty in AI governance.

Peng Chengxin, dean and special professor of Kaiyuan Law School of Shanghai Jiaotong University, said in his speech: The fundamental reason for ethical governance in a digital society is people. The credibility and goodness of corporate ethics and technology is ultimately the credibility and goodness of people. Therefore, it is necessary to improve legal supervision, clarify basic legal responsibilities, and establish people's behavioral norms.

In order to further regulate the role of "people" in the governance of science and technology ethics, the "Opinions on Strengthening the Governance of Science and Technology Ethics" issued in March this year requires institutions engaged in scientific and technological activities such as life sciences, medicine, and artificial intelligence, and the research content involves science and technology. In ethical-sensitive fields, a science and technology ethics (review) committee should be established.

At present, many large domestic enterprises have carried out in-depth work on artificial intelligence ethics and governance.

At the 2022 World Artificial Intelligence Conference, Alibaba Group announced the establishment of a Science and Technology Ethics Governance Committee and introduced 7 external consultants to strengthen tripartite supervision. The committee proposed six major scientific and technological ethical principles: "people-oriented, universal integrity, safety and reliability, privacy protection, credibility and control, open and co-governance", focusing on promoting algorithmic governance, privacy protection, ethical system construction, etc., exploring relevant cutting-edge technologies, and improving The company's overall technological introspection and self-discipline.

SenseTime coordinated the construction of the work system for the ethical governance of artificial intelligence in January 2020, and released the "Balanced Development" Artificial Intelligence Governance White Paper - SenseTime Artificial Intelligence Ethics and Governance Annual Report (2022)" this year. The report points out that artificial intelligence governance has entered the stage of implementation practice mainly based on technology verification. Based on this, SenseTime Technology further proposed the development of "responsible and evaluable" artificial intelligence based on the "balanced development" of artificial intelligence ethics last year, and regarded it as the vision and goal of SenseTime's artificial intelligence governance. Create a closed loop of ethical governance.