在“人工智能+”背景下，ChatGPT、Sora及文心一言等生成式人工智能应用（AIGC）横空出世，显著提升了人类的生产效能与创新潜能，但亦伴生伦理价值失衡、伦理规范失控、伦理关系失调等风险。基于技术风险维度分析，算法的隐秘性与智能技术的不可预测性、价值理性与工具理性的内在张力、人类风险感知与应对机制的局限性，是AIGC伦理风险的核心成因。鉴此，AIGC伦理风险管控正趋于“可信治理”范式，其核心要旨在于确保技术的可控性、可问责性、公平性、可靠性、可解释性和安全性。在可信治理引导下，应完善数据风险治理架构，优化侵权风险防控与责任追究体系，健全伦理治理制度，以赋能新质生产力发展。

At the moment when AI technology is fully penetrated into all fields of social life, the rise of AIGC(Artificial Intelligence Generated Content) represented by ChatGPT, Sora and ERNIE Bot marks the leap of AI technology to a new stage of development, and indicates the approach of the era of "artificial general intelligence". However, while AIGC greatly amplifies human innovation potential, it also presents considerable challenges to the existing ethical governance system. Against this backdrop, this article examines the current state and future trends of AIGC in China, analyzing its ethical risks and generative logic. Drawing from the ethical risk governance trends of AIGC globally, the article proposes a governance framework that is not only attuned to China's national context but also forward-looking. This is aimed at providing robust theoretical guidance and practical paths for the standardized development of AIGC.
This study commences with an analysis of the potential ethical risks associated with AIGC, considering its technical attributes. Initially, it addresses the risk of ethical value imbalance, primarily evident in the intensification of algorithmic discrimination. Following this, it examines the risk of ethical norms being beyond control, which is predominantly showcased in the challenge of accountability. Lastly, it explores the risk of ethical relationship imbalance, which is chiefly characterized by a diminution of human agency. From the perspective of technological risk, the root cause of AIGC ethical risk lies in the complexity and uncertainty of technology. Particularly under the 'black box' phenomenon of algorithms, the behavior of AIGC becomes challenging to anticipate and manage, thereby amplifying ethical risks. At the same time, the tension between technological rationality and value rationality, as well as the limitations of human risk perception and response, further deepen the complexity of ethical risks.
Considering the trends in AIGC governance, this article advocates "trusted governance" as the core strategy to mitigate AIGC ethical risks. It underscores the importance of technology being controllable, accountable, fair, reliable, interpretable, and secure, with the goal of ensuring that technological advancements are transparent, equitable, and contribute positively to society. At the same time, the Artificial Intelligence Law (scholars' proposal draft) also provides important governance basis and guides the legal path of AIGC trusted governance.
Under the principle of 'trusted governance, this article proposes three core strategies. First, it calls for the establishment of a robust data risk governance framework. By integrating cross-border collaborative governance and internal fine governance, it emphasizes adherence to 'reliability' standards. This includes enhancing regulations for cross-border data flows and promoting data ethics and compliance within enterprises to ensure data usage is reliable and secure. Second, it advocates for an optimized liability attribution mechanism to address AIGC infringement risks. This involves assessing subject obligations, product defects, and infringement liability in light of AIGC's technical characteristics to enforce the 'accountability' standard. Third, it recommends integrating a 'people-oriented' approach into the AIGC ethical governance system. This aims to find technological solutions that are tailored to China's context and capable of addressing its unique challenges. In practice, it shall involve two aspects: first, at the organizational level, creating a dedicated AIGC ethical governance body to oversee ethical governance and supervision, thereby enhancing AIGC's controllability; and second, at the normative level, harnessing the 'complementary advantages' of policies, laws, and technology to uphold standards of 'fairness,' 'safety,' and 'interpretability,' guiding AIGC towards positive development.
In the future, it is essential to further integrate market development with China's national conditions, adopting 'trusted governance' as the core and the 'Proposal Draft' as the foundation to construct a comprehensive ethical risk governance system for AIGC. Moreover, there is a need for ongoing exploration in the realm of ethical governance, focusing on the organic integration of technological ethics with legal governance. This approach aims to devise AI ethical governance strategies that are distinctively Chinese, thereby equipping to adapt to and potentially spearhead the high-quality development of the new round of the global digital economy.