AI Ethics

Global Digital Management|Two Key Consensuses On The First Global Agreement On Artificial Intelligence Ethics

Global Digital Management|Two Key Consensuses On The First Global Agreement On Artificial Intelligence Ethics

Global Digital Management|Two Key Consensuses On The First Global Agreement On Artificial Intelligence Ethics

Compiled by Pengpai News reporter Lu Na. Here is the weekly report of the

Compiled by Pengpai News reporter Lu Na

Here is the weekly report of the "Global Digital Governance" column, tracking recent global digital governance trends

On November 24, UNESCO () adopted the first global agreement on artificial intelligence ethics, "Recommendation on the Ethics of Artificial Intelligence" (hereinafter referred to as the "Recommendation") at the 41st General Assembly, for 193 member states. use. The Recommendation defines common values ​​and principles on artificial intelligence technology and application, and is used to guide the establishment of necessary legal frameworks, ensure the healthy development of artificial intelligence, promote the technology to serve human beings, society, environment and ecosystems, and Prevent potential risks.

As a key technological innovation leading the fourth scientific and technological revolution, artificial intelligence has brought significant and far-reaching positive impacts on global social and economic development, such as increasing convenience of life and people's livelihood, improving government and enterprise operational efficiency, helping to respond to climate change and Poverty and hunger problems. However, the technology also poses unprecedented challenges, including major threats to personal privacy and dignity, intensified gender and racial bias, and a surge in large-scale surveillance risks.

Given that this technology is both creative and destructive in its actual application process, but it has long lacked global governance standards. In 2018, it launched the "Building an Ethical Framework for the Use of Artificial Intelligence for the World" project, and selected 24 from all over the world. The experts jointly wrote and completed the "Artificial Intelligence Ethics Recommendation" for three years, and after more than 100 hours of multilateral negotiations and repeated revisions between 193 member states, it was finally approved in November this year. This is not only the world's first rule framework and common program for artificial intelligence ethics, but also the broadest consensus reached at the level of governments around the world, providing strong support for the next step of formulating specific international regulations and technical standards to regulate the development of artificial intelligence. , which can be regarded as another victory of multilateralism.

The ethical framework of artificial intelligence established in the Recommendation is mainly composed of three parts: values, ethical principles and policy guidance. The text content is very comprehensive and focuses on the balance of multi-party concepts and interests. Among them, the values ​​of artificial intelligence emphasize: 1. Respect, protect and promote human rights, basic freedoms and human dignity; 2. Protect the vigorous development of the environment and ecosystems; 3. Ensure diversity and inclusion; 4. In peace and justice Resurrection with an interconnected society. Ethical principles mainly include: proportionality and non-damage, security, fairness and non-discrimination, sustainability, privacy and data protection, human supervision and decision-making, transparency and interpretability, responsibility and accountability, technical awareness and literacy 10 aspects including multi-stakeholder collaborative governance. Policy guidance concerns ethical impact assessment, ethical governance and management, data policy, development and international cooperation, environment and ecosystems, gender, culture, education and research, communication and information, economy and labor, health and social well-being, monitoring and assessment There are 12 subdivisions in total. At the same time, the Recommendation encourages all member states to consider the addition of independent AI ethics officials or other relevant mechanisms to monitor the impact of the application of the technology.

Based on comprehensive consideration and extensive coverage, the most core content of the Recommendation focuses on calling on countries to take data protection actions, prohibiting social scoring and large-scale monitoring, monitoring and evaluating the social and environmental impacts of artificial intelligence systems. Moreover, the Recommendation does not blindly follow the existing standards of Western artificial intelligence technology-first countries on these issues, but through regional expert consultation meetings, intergovernmental expert consultations, and solicitation of opinions from various member states during the compilation process. Come to widely listen to and adopt the opinions and suggestions of developing countries such as China, strive to balance value differences, and promote the formation of multilateral consensus. Among them, two consensuses are particularly critical.

First, the Recommendation determines "sustainable development" as the overall vision and important way for the development of global artificial intelligence, and advocates that the development of artificial intelligence technology and applications are prioritized to facilitate the realization of the sustainable development goals, and requires close attention and Prevent all kinds of risks that hinder the realization of this goal. At the same time, the "Proposal" is not limited to the debate about whether development should be first or governance should be first, emphasizing that the two are coordinated so that artificial intelligence systems can empower harmony between humans, society and ecology throughout their life cycle. Symbiosis.

Secondly, in terms of innovation in artificial intelligence ethics and governance paradigm, the "Recommendation" advocates the path choice of "coordinated governance". The Recommendation emphasizes that ethical challenges brought by artificial intelligence applications such as face recognition, automated decision-making, and social scoring cannot solve problems only through prohibiting use, but rather, ethical standards based on multilateral consensus should be integrated into artificial intelligence systems. All processes of design, research and development, deployment and use are adapted to corresponding governance rules and means to prevent problems before they occur. At the same time, artificial intelligence governance cannot be effectively achieved through government or intergovernmental cooperation alone, but depends on the in-depth coordination of multiple parties in the entire technology ecosystem. Stakeholders such as government, enterprises, R&D, users, academia, and media all need to assume corresponding responsibilities and carry out full-process collaborative governance.

Through relatively equal and extensive public discussion, the establishment of the Recommendation and the first global ethical framework for artificial intelligence has been facilitated. This not only brings together 193 countries around the world to have a consensus on the sustainable development of this emerging technology, but also further opens up new prospects for future multilateral collaborative governance. In the next step, it is also necessary to implement the values, principles and policies in it, promote countries to form enforceable standards and rules, turn ideas into actions, and integrate ideals into reality.

rule

The United States and Europe launch a joint dialogue on technology competition policy

On December 7, European Commission Executive Vice President Margaret Vestager, FTC Chairman Lena Khan and U.S. Department of Justice’s Assistant Attorney General Jonathan Kant launched the U.S.-EU joint effort in Washington. Respond to the technical competition policy dialogue and issue a joint statement to strengthen policy coordination and joint law enforcement between the two sides in the rapidly developing technology field. The statement pointed out that the United States and Europe adhere to common democratic values ​​and belief in healthy market competition, and face common challenges such as network effects, massive data, and interoperability in the field of digital competition law enforcement. Therefore, the United States and Europe will further strengthen coordination in policy and law enforcement. (Source: European Commission official website)

UK releases groundbreaking standards for algorithm transparency

On November 29, the UK Central Digital Office (CDDO) released the algorithm transparency standard, aiming to provide government departments and public institutions with clear information on algorithmic tools used to support decision-making, fulfilling its presence in the National Data Strategy and National Artificial Intelligence Strategy. Commitment made. Algorithm transparency standards cover transparency data standards, transparency templates and action guides, etc., which will help organizations improve the transparency of algorithm tools. (Source: UK government official website)

Myanmar Rohingya refugees sued for violence, demanding $150 billion in compensation

On December 6, a Rohingya refugee group in Myanmar filed a class action lawsuit in the United States, accusing social media platforms of promoting violence against persecuted minorities, demanding $150 billion in damages. It said that although misinformation and hate speech were not prevented in a timely manner before, remedial measures have been taken afterwards, including banning the use of and the military after the February 1 coup in Myanmar. It has stated that it is protected by "Section 230" of the US Internet Law and is exempt from liability for content posted by users. Rohingya refugee groups said they would seek to make a claim through Myanmar law if "Article 230" is proposed as a defense. (Source: Reuters)

policy

The U.S. Department of Defense plans to reorganize three key technology offices

On December 1, the U.S. Department of Defense announced that it will reorganize the Department of Defense Digital Services, the Joint Artificial Intelligence Center and the Chief Data Office into the Office of the Chief Data and Artificial Intelligence Office to simplify the process and create more information for the use of artificial intelligence and data. A cohesive method. Under the proposed plan, the three offices remain independent, but all report to the new office. The move will provide three offices with a clearer organizational structure, helping them master more data and apply it to artificial intelligence, thereby promoting joint command and control efforts across the region. (Source: American Association for Foreign Relations)

The Ministry of Industry and Information Technology predicts that China's big data industry will exceed 3 trillion yuan in 2025

On November 30, the Ministry of Industry and Information Technology held a press conference to introduce the relevant situation of the "14th Five-Year Plan for Big Data Industry Development". The "Plan" proposes that by 2025, the scale of my country's big data industry is expected to exceed 3 trillion yuan, with an average annual compound growth rate of about 25%, and a basic modern big data industry system with strong innovation, high added value and independent control. form. The "Plan" sets key tasks on six aspects: accelerating the cultivation of the data element market, giving full play to the characteristics and advantages of big data, consolidating the foundation of industrial development, building a stable and efficient industrial chain, creating a prosperous and orderly industrial ecology, and building a solid data security line of defense. Six special actions have been arranged, including improvement of data governance capabilities, development and application promotion of key standards, improvement of industrial big data value, development and utilization of industry big data, leap in corporate entity development level, and data security shield casting. In addition, the "Plan" also proposes to accelerate the construction of a national integrated big data center system, promote the construction of the national industrial Internet big data center, guide the integrated development of large, medium and small enterprises and collaborative innovation of upstream and downstream of the industrial chain, and support traditional enterprises to carry out the divestiture and reorganization of big data business. (Source: China Government Website)

Supervision

FTC filed lawsuit against Nvidia's acquisition of ARM

On December 2, the Federal Trade Commission (FTC) announced that it had filed a lawsuit on the grounds of antitrust to prevent Nvidia from acquiring ARM for $40 billion. The FTC believes that the deal will kill market innovation in various technologies such as cloud computing, data centers and autonomous vehicles, and will "unfairly weaken" Nvidia's competitors. The lawsuit is another regulatory blow to Nvidia's ARM acquisition. Previously, the UK government and the European Commission had begun in-depth antitrust investigations into the deal. (Source: Reuters)

Industry and Technology

Meta and Amazon AWS reach cloud service cooperation agreement

On December 1, Meta announced a partnership with Amazon Cloud Services (AWS) to use AWS as its long-term strategic cloud service provider to expand the use of AWS compute, storage, database and security services to supplement Meta’s current Some local infrastructure. Meta will also use AWS's computing services to accelerate the R&D of its AI team and will run third-party cooperative applications on the AWS platform. (Source: Amazon official website)

Urban digitalization

The State Council issued a document to support Beijing Urban Sub-Center to vigorously develop the digital economy and promote the construction of smart cities

On November 26, the State Council issued opinions on supporting the high-quality development of Beijing's urban sub-center. The opinions propose that by 2025, the functions of urban sub-center green cities, forest cities, sponge cities, smart cities, humanistic cities and livable cities will be basically formed. The opinions are clear, strengthen the leadership of scientific and technological innovation, focus on the fields of new generation information technology, intelligent manufacturing, and other fields, implement a number of major national scientific and technological projects and application demonstration projects, guide the layout of innovation chains and industrial chains in urban sub-centers and surrounding areas, and vigorously develop the digital economy . Focusing on the fifth-generation mobile communications (5G) network, artificial intelligence, cloud computing, big data, Internet protocol version 6 (IPv6), etc., we will build a number of demonstration application new applications in the fields of smart cities and digital rural construction. Scenarios, support the pilot project of sandboxes for scientific and technological application scenarios. (Source: China Government Website)

Baltimore, the city of U.S. plans to build a universal and intelligent network to narrow the digital divide

On November 30, the mayor of Baltimore, the mayor of Baltimore, announced that it would allocate $35 million to provide network access relief to communities and residents most affected by public health emergencies. Among them, the first $6 million will be used to significantly expand public Internet access to narrow the digital divide in the city. Given that Internet access has become the most critical and basic public infrastructure, achieving network access fairness has become one of the key connotations of future smart cities. Baltimore plans to continue to invest in the construction of open and inclusive intelligent network infrastructure throughout the city over the next decade to ensure that every citizen can access the Internet equally and conveniently. (Source: Baltimore City Government Official Website)

More