导论：AI时代的伦理紧迫性 / Introduction: The Ethical Imperative in the Age of AI

当2016年AlphaGo在围棋对弈中击败李世石九段的那一刻，世界见证了一个历史性的转折。这场胜利不仅展示了深度学习技术的惊人能力，更预示着人工智能（Artificial Intelligence, AI）将以前所未有的深度和广度渗透人类社会的每一个角落。然而，正当公众为技术进步欢欣鼓舞之际，ProPublica记者Jackie Mader和她的团队在同年发布的一项调查，却揭示了AI发展背后一个令人不安的真相：那些被吹捧为客观公正的算法，实际上可能在系统性地延续和放大人类社会中的偏见与歧视。这一发现如同一面镜子，映照出AI技术光鲜外表下隐藏的伦理困境，也标志着AI伦理从一个抽象的哲学议题转变为一个紧迫的现实挑战。

The moment AlphaGo defeated Lee Sedol, a professional 9-dan Go player, in 2016, the world witnessed a historic turning point. This victory not only demonstrated the astonishing capabilities of deep learning technology but also presaged that artificial intelligence would penetrate every corner of human society with unprecedented depth and breadth. However, while the public celebrated this technological triumph, an investigation published that same year by ProPublica journalist Jackie Mader and her team revealed an unsettling truth lurking behind the gleaming facade of AI development: the algorithms being hailed as objective and impartial were, in fact, systematically perpetuating and amplifying the biases and discrimination embedded in human society. This discovery functioned like a mirror, reflecting the ethical dilemmas hidden beneath the glamorous surface of AI technology, and marked the transformation of AI ethics from an abstract philosophical discourse into an urgent practical challenge.

本章目标与结构概述 / Chapter Objectives and Structure Overview

本章作为全书的开篇，承担着建立分析框架、厘清核心概念、为后续深入讨论奠定基础的多重使命。我们将从三个维度展开论述：首先，系统界定人工智能伦理学的学科边界，澄清AI Ethics、Responsible AI、AI Governance等密切相关但本质不同的概念范畴；其次，通过一系列具有代表性的真实案例，深入剖析AI系统伦理失误的严重后果及其深层成因；最后，隆重推出本书采用的伦理分析工具箱，包括规范伦理学方法、价值敏感设计、伦理考察流程以及跨学科分析框架，为后续章节的展开提供方法论支撑。

This opening chapter, serving as the foundation of the entire book, carries the multiple missions of establishing an analytical framework, clarifying core concepts, and laying the groundwork for in-depth discussions to follow. We will develop our arguments from three dimensions: First, systematically delineating the disciplinary boundaries of AI ethics and clarifying the related yet fundamentally distinct conceptual categories of AI Ethics, Responsible AI, and AI Governance; second, through a series of representative real-world cases, deeply analyzing the serious consequences of ethical failures in AI systems and their underlying causes; third, formally introducing the ethical toolkit employed throughout this book, including normative ethics approaches, Value-Sensitive Design, ethical review processes, and interdisciplinary analytical frameworks, to provide methodological support for subsequent chapters.

人工智能伦理的学科边界：概念辨析与范畴厘清 / The Disciplinary Boundaries of AI Ethics: Conceptual Clarification and Categorical Delineation

从信息伦理到机器道德：学科演进的历史脉络 / From Information Ethics to Machine Morality: The Historical Trajectory of Disciplinary Evolution

人工智能伦理学的诞生并非凭空出现，而是植根于一个更为悠久的思想传统——信息伦理（Information Ethics）。1985年，著名计算机科学家和伦理学家Deborah G. Johnson在其开创性论文"Ethics and Computing: Living Responsibly in a Computerized World"中，首次系统性地将伦理问题引入计算机科学领域，标志着计算伦理（Computer Ethics）作为一个独立研究领域的正式确立。这一时期的研究主要关注计算机技术本身带来的伦理问题，包括软件知识产权、计算机犯罪、隐私保护等议题。

The emergence of AI ethics did not arise spontaneously but was rooted in a more venerable intellectual tradition—information ethics. In 1985, the renowned computer scientist and ethicist Deborah G. Johnson, in her groundbreaking paper "Ethics and Computing: Living Responsibly in a Computerized World," systematically introduced ethical issues into the field of computer science, marking the formal establishment of computer ethics as an independent research domain. During this period, research primarily focused on ethical problems inherent in computer technology itself, including software intellectual property, computer crime, privacy protection, and related issues.

随着互联网的普及和信息技术的深化发展，信息伦理的研究范畴逐步扩展为网络伦理（Cyber Ethics），开始关注网络空间中的行为规范、虚拟身份、数字鸿沟等问题。进入21世纪后，大数据、云计算、物联网等新兴技术的兴起，使得信息伦理的研究对象从单纯的信息处理扩展到更为广泛的数据治理（Data Governance）领域。Floridi教授在其2013年出版的《信息伦理学》（The Ethics of Information）中，提出了著名的"信息哲学"（Philosophy of Information）框架，将信息视为理解现实世界的基础性概念，为人工智能伦理的兴起奠定了哲学基础。

With the proliferation of the internet and the deepening development of information technology, the scope of information ethics gradually expanded into cyber ethics, beginning to address behavioral norms in cyberspace, virtual identities, the digital divide, and related issues. After entering the 21st century, the rise of emerging technologies such as big data, cloud computing, and the Internet of Things expanded the research objects of information ethics from pure information processing to the broader field of data governance. Professor Luciano Floridi, in his seminal 2013 work "The Ethics of Information," proposed the influential "philosophy of information" framework, treating information as a fundamental concept for understanding the real world and laying the philosophical groundwork for the emergence of AI ethics.

人工智能伦理学的正式形成，可以追溯到2015年前后。这一时期，深度学习技术的突破性进展使得AI系统在图像识别、自然语言处理、语音识别等领域达到或超越人类水平，AI开始从实验室走向大规模商业应用。与此同时，算法歧视、数据泄露、自动化失业等社会问题日益凸显，引发了学术界、产业界和公众的广泛关注。麻省理工学院的Patrick Lin博士、斯坦福大学的Jerry Kaplan博士、牛津大学的Nick Bostrom教授等学者，成为这一新兴领域的先驱研究者，他们的工作为人工智能伦理学的学科建制化做出了重要贡献。

The formal formation of AI ethics can be traced to around 2015. During this period, breakthrough advances in deep learning technology enabled AI systems to reach or surpass human-level performance in image recognition, natural language processing, speech recognition, and other domains, marking AI's transition from laboratory settings to large-scale commercial applications. Simultaneously, social issues such as algorithmic discrimination, data breaches, and automated unemployment became increasingly prominent, sparking widespread concern among academia, industry, and the public. Scholars such as Dr. Patrick Lin at MIT, Dr. Jerry Kaplan at Stanford, and Professor Nick Bostrom at Oxford emerged as pioneering researchers in this new field, making significant contributions to the institutionalization of AI ethics as an academic discipline.

核心概念辨析：AI Ethics、Responsible AI与AI Governance / Core Conceptual Clarification: AI Ethics, Responsible AI, and AI Governance

尽管"人工智能伦理"这一术语在公共讨论中被广泛使用，但其具体含义往往模糊不清，甚至被随意替换为其他相关概念。为了建立本书的分析基础，我们必须首先澄清三个核心概念的边界与关联：AI Ethics（人工智能伦理）、Responsible AI（负责任人工智能）和AI Governance（人工智能治理）。

Although the term "AI ethics" is widely used in public discourse, its specific meaning is often ambiguous and is even arbitrarily interchanged with related concepts. To establish the analytical foundation for this book, we must first clarify the boundaries and interrelationships of three core concepts: AI Ethics, Responsible AI, and AI Governance.

AI Ethics（人工智能伦理）是人工智能伦理学的学术称谓，其研究对象是AI系统本身固有的伦理问题，即AI系统在设计、开发、部署和使用过程中所涉及的道德价值、规范和原则问题。从哲学传统来看，AI Ethics主要属于应用伦理学（Applied Ethics）的范畴，借鉴并发展了生命伦理学（Bioethics）、环境伦理学（Environmental Ethics）等相邻领域的理论资源。AI Ethics的核心问题包括：机器是否能够具有道德主体性（moral agency）？AI系统是否应该承担道德责任？如何确保AI系统的决策符合人类道德规范？AI的发展是否会对人类尊严构成威胁？

AI Ethics is the academic designation for the study of AI ethics, with its research objects being the inherent ethical issues of AI systems themselves—the moral values, norms, and principles involved in the design, development, deployment, and use of AI systems. From the perspective of philosophical tradition, AI Ethics primarily belongs to the category of applied ethics, drawing upon and developing theoretical resources from neighboring fields such as bioethics and environmental ethics. Core questions in AI Ethics include: Can machines possess moral agency? Should AI systems bear moral responsibility? How can we ensure that AI system decisions conform to human moral norms? Does the development of AI pose threats to human dignity?

Responsible AI（负责任人工智能）这一概念更强调实践导向，它关注的是如何在技术开发的具体实践中落实伦理原则，确保AI系统的开发和使用符合道德规范。Responsible AI的概念框架通常包括公平性（Fairness）、透明性（Transparency）、可解释性（Explainability）、隐私保护（Privacy）、安全性（Security）、鲁棒性（Robustness）等技术维度。从历史渊源来看，"负责任创新"（Responsible Innovation）的概念最早由欧盟委员会在2011年发布的《负责任创新路线图》（Roadmap on Responsible Innovation）中提出，随后被引入AI领域，成为指导AI开发实践的重要理念。