人工智能科技倫理審查制度的體系化建構

Model Classification and System Connection of Artificial Intelligence Science and Technology Ethics Review System

目前我國科技倫理審查實踐存在審查形式化、虛置化等現實問題，並且科技倫理治理固有的“軟治理”特徵也使得該制度難以實現預期的制度目標。實踐中倫理失控風險主要表現為既有的科技倫理審查制度難以有效控制人工智能科技倫理風險，造成這一現象的根源在於未能明確人工智能科技倫理審查制度的功能定位，導致在審查範圍、審查標準、審查程序等方面存在模糊性問題。該審查制度雖然屬於典型的風險治理範式，但其以非強制性的社會倫理規範為制度基礎，審查結果屬於倫理道德層面的判斷，其性質屬於科技創新活動的合理性審查。基於此種定位，應當適當區分科技倫理委員會與科技倫理審查委員會的定位差異，僅限於涉及敏感研究領域才強制要求被審查對象組建科技倫理審查委員會，並將審查結果與安全風險評估結果一併作為採取特定治理措施的風險依據。

At present, there are practical problems such as formalization and virtualization in the practice of science and technology ethics review in my country, and the inherent ''soft governance'' characteristics of science and technology ethics governance also make it difficult for the system to achieve the expected institutional goals. In practice, it is manifested that the existing science and technology ethics review system is difficult to effectively control the ''out of control'' phenomenon of artificial intelligence science and technology ethics risks. The root cause of this phenomenon is the failure to accurately understand the functional positioning of the AI science and technology ethics review system, and there are ambiguities in the scope of review, review standards, and review mechanisms. First, the operation mechanism of the Science and Technology Ethics Review Committee and its own legal nature are always in doubt. Second, there is no consensus on how to improve the governance effect of the ''weak constraint'' effect of science and technology ethics review through the rule of law. Third, whether the formalization of science and technology ethics review can be solved through legal responsibility is worth discussing. Although this mechanism is a typical risk governance mechanism, it is still based on non-mandatory social ethical norms. The review results belong to the judgment of the ethical level, and it is more appropriate to understand it as a rationality review of scientific and technological innovation activities. In terms of risk identification, the realization of the functions of the artificial intelligence science and technology ethics review system needs not only to be improved in terms of review items and review processes, but also to be supplemented in the feedback and rectification of review results. In terms of risk prevention, the artificial intelligence science and technology ethics review system can target unexpected artificial intelligence science and technology innovation activities and discover technical application methods that may violate ethics and damage personal dignity in the early stage. In terms of rationality judgment, the current artificial intelligence science and technology ethics review mainly solves the potential science and technology ethics risks at the technical application level and judges whether the corresponding technical application methods are reasonable. Based on this positioning, the positioning differences between the Science and Technology Ethics Committee and the Science and Technology Ethics Review Committee should be appropriately distinguished. Only when sensitive research fields are involved, the subject of the review is required to form a Science and Technology Ethics Review Committee, and the review results and security risk assessment results are used as the risk basis for taking specific governance measures. The application of science and technology ethics review should involve three situations. The first situation involves new ways of applying artificial intelligence technology; the second situation involves scientific and technological innovation activities that involve significant abnormal scientific and technological ethical risks; and the third situation involves gatekeeping. Scientific and technological innovation and application activities engaged in by high-tech enterprises with human status.