On October 17, 2024, the world-famous Nobel Prize finally announced its list for the new year. The winners shocked the global scientific community: the Nobel Prize winners in physics and chemistry are closely related to artificial intelligence. This intuitively reveals the key position of artificial intelligence in modern scientific research and further triggers extensive discussions about artificial intelligence.

Entering a new stage of artificial intelligence: the origin of the social ethics governance crisis

Since the concept of "artificial intelligence" was first formally proposed at the Dartmouth Conference in 1956, artificial intelligence has traveled thousands of miles and spanned two "cold winter periods." With the breakthrough of technology, the reduction of cost and the popularization of application, in 2016, the artificial intelligence developed by Google defeated the human Go player Lee Sedol in the human-machine battle, and artificial intelligence returned to the public eye. The rapid development of the Internet has pushed mankind into the era of big data. The three elements of data, algorithms, and computing power have advanced hand in hand. Deep learning technology, mainly deep neural networks, has begun to rise and continues to make breakthroughs. Artificial intelligence has gradually "flyed into the homes of ordinary people." In 2022, Dall-E2, , and other applications will come one after another, and the emergence of Dall-E2 will make the scenery of generative artificial intelligence unprecedented. The emergence of generative artificial intelligence marks a new era of artificial intelligence, and at the same time it also triggers in-depth discussions on the risks of artificial intelligence, especially its impact on social ethics.

The three books included in "Unicorn·Artificial Intelligence" (Seventh Series) are closely related to this. "Artificial Intelligence Liability" edited by Sebastian Lose and others is a collection of papers from the 7th Münster EU Law and Digital Economy Symposium. It closely follows the EU's "Artificial Intelligence Liability Directive" proposal and the "Product Liability Directive" revised proposal. It gathers wisdom from four perspectives: fault liability and strict liability, producer liability and operator liability, causality and fault and burden of proof, recourse and insurance, and systematically sorts out the important innovations contained in the liability system in the two proposals and the corresponding amendment suggestions. "It's Private, It's Political" written by Alice E. Marwick focuses on women and marginalized groups in society, and provides an in-depth analysis of the issue of privacy inequality in the digital age. By constructing a novel "online privacy" system, it reveals the shortcomings of the current privacy protection system and highlights the practical need to achieve privacy protection from a collective political level. "Medical Confidentiality and Privacy" co-authored by Thierry van Swievelt and Nicola Glover-Thomas focuses on the two core areas of privacy and medical confidentiality. It displays the diverse methods adopted by different countries and regions in handling sensitive patient information from a global perspective. It provides a reference for practitioners in the medical industry to fully understand their own responsibilities and promotes mutual learning among countries.

Grasping the new rules of artificial intelligence: exploring paths for social ethical governance

Based on the autonomy and opacity of artificial intelligence, social ethics in the era of artificial intelligence are facing new challenges, one of which is the identification and allocation of tort liability. The book "Responsibility in Artificial Intelligence" can be described as a vivid footnote on the EU's response to this challenge.

This book has three distinctive features: First, it reiterates the position of the EU on the subject status of artificial intelligence, that is, in the context of the evolving EU tort law, the artificial intelligence system itself does not have legal subject status, and therefore cannot directly be the bearer of tort liability. This results in the attribution of responsibility needing to be realized through other subjects, such as algorithm designers, data providers and users, etc. Second, although the authors of each paper are different, the arrangement of this book is clear and logically organized. As the outcome of the EU Law and Digital Economy Seminar, this book closely follows the related concepts, adjustment objects and interrelationships of the two proposals of the Artificial Intelligence Liability Directive and the Product Liability Directive as basic topics, and clearly puts forward core propositions such as the increasing necessity for the supervision of artificial intelligence and, by extension, the need for further development of EU law in the field of tort law. Around these claims, the basic argument is divided into four steps: first, study the liability system itself from a general level; second, explore the core aspects of different liability systems, namely product liability and operator liability; third, focus on the difficulties faced by victims in bearing the burden of proof; and finally, discuss the issue of allocating damages among different subjects. Third, this book combines specific scenarios where artificial intelligence is applicable in real life, providing a collision of ideas on the consistency of relevant norms in this field and the balance of burdens between responsible persons and potential victims. Focusing on two important proposals, this book adopts a comparative law perspective and examines the different subfields or "islands" of EU law involved in the field of artificial intelligence. For example, Christian Windhorst believes that the few gaps left by the "Product Liability Directive" proposal can only be solved by supplementing the liability of alternative operators; while Jorge Borges proposed that the "Artificial Intelligence Liability Directive" proposal cannot fill all the gaps between the responsibilities arising from the autonomy and opacity of artificial intelligence systems, that is, the legacy of the "Product Liability Directive" proposal. Additional strict liability systems may still be needed to deal with the liability gaps caused by the autonomous operation of artificial intelligence systems.

The value contained in this book is self-evident. However, as mentioned in the book, legislation in the field of artificial intelligence is currently fragmented, so the main task for legal scholars and practitioners remains to try to find uniformity in the ever-expanding tort liability law. However, based on the various papers in this article, it seems that this goal has not been fully achieved, or even difficult to achieve. To be more specific: the question raised at the meeting - "causality" plays an important role within the normative scope of the directive proposal. But should this concept be, or is it possible to be, applied consistently in EU law? Thinking further, if the concept of "causation" should be harmonized, should it be applied within the relevant subfield of EU law, or should it be applied more broadly within the EU legal system? Similarly, can the concept of "fault" mentioned in the proposal for the "Artificial Intelligence Liability Directive" only be applicable within the scope of the directive, or is there the possibility of application outside the scope? For the purpose of EU legal consistency, people may also need to consider the relationship between these rules and corresponding rules in other areas of EU law, especially the issue of consistency of protection standards. In addition, whether it is product liability or operator liability, under what circumstances the liability should be attributed to the manufacturer or producer, there is still a lot of confusion after discussion.

As the editor said: "Artificial intelligence connects all areas of life, connects social individuals and public interests, and forms a multi-dimensional pattern in which 'one hair affects the whole body'." The other two books are closely related to the overall gist of "Artificial Intelligence Responsibility", but their focus is different, that is, they focus on privacy issues and artificial intelligence infringement in specific fields for analysis and research. "It's a Private Matter, It's Political" is a network-based artificial intelligence work that focuses on the social field. What makes it stand out is that it breaks away from the inherent connection between privacy and individuals under the big privacy rights system, and turns to explore the socialization of privacy. It studies the harm of privacy infringement to different groups from a political perspective, and the power differences at the processing level. "Medical Confidentiality and Privacy" focuses on the trust relationship in the medical field. By collecting a large number of policies and practices in various countries and conducting comparative analysis, it reveals the diverse practices of medical privacy protection and provides rich practical resources for the academic and practical circles.

Drawing up a new vision for artificial intelligence: Prospects for the future of social ethical governance

Mill mentioned in "On Liberty": "No matter how correct an opinion is, or even an objective truth, if it cannot withstand sufficient and fearless discussion from time to time, it can only become a dead dogma rather than a living truth." Academic progress requires continuous dialogue, and the deepening of dialogue requires seeking common ground while reserving differences, so as to achieve a "spiral upward" in the level of thought.

At present, my country's artificial intelligence legislation has carried out many useful explorations. In theory, documents such as the "Artificial Intelligence Law of the People's Republic of China (Draft of Scholars' Suggestions)" and "Key Institutional Suggestions on Artificial Intelligence Legislation" were released. In practice, a multi-level, regional and domain-based legal governance framework has been formed. For example, the "New Generation Artificial Intelligence Development Plan" as a macro-guidance document, the "Interim Measures for the Management of Generative Artificial Intelligence Services" which specifically regulates generative artificial intelligence, the country's first local regulation on the artificial intelligence industry, the "Shenzhen Special Economic Zone Artificial Intelligence Industry Promotion Regulations", etc. However, a unified and systematic set of specialized laws and regulations has not yet been formed. How should laws with natural lag respond to the ever-changing artificial intelligence technology? When the legal rule system has not yet been perfected, where can my country's social ethics governance go in the era of artificial intelligence? Countless new problems and challenges await legal practitioners to jointly address.

This series of books aims to focus on digital transformation and artificial intelligence, and grasp and realize the unification and systematization of artificial intelligence governance. This has strong reference value and reference significance for my country's artificial intelligence legislation and theoretical research. It needs to be emphasized that these three books are not "all-in-one" and cannot give us comprehensive responses and answers. However, the rich connotations contained in them may be able to open a door to the best answer while bringing us the collision of wisdom crystallization and reflection on future rules. As editor-in-chief Professor Peng Chengxin said in the preface: "We look forward to a more just, safe and reliable era of artificial intelligence, an era in which law, technology and life promote each other and make progress together, and an era in which human wisdom and artificial intelligence coexist harmoniously."

""Unicorn·Artificial Intelligence" Volume 7"

Editor-in-Chief Peng Chengcheng

Published by Shanghai People's Publishing House in 2024