构建稳健敏捷的人工智能伦理与治理框架

Building a Robust and Agile Framework for AI Ethics and Governance

中文关键词: 人工智能伦理与治理 人工智能信任机制 “技术- 伦理”矫正 敏捷治理

英文关键词: artificial intelligence ethics and governance  trust mechanism for AI  “technicalethical” correction,agile governance

中文摘要:

      在各国和地区纷纷提出人工智能发展规划和战略的同时，人工智能的伦理风险与治理也成为全球共 同关注的焦点。鉴于不同区域社会环境和文化背景的不同，很难构建单一的全球性人工智能伦理与治理框架。 中国在总体上对于人工智能为经济、社会、企业和个人福祉带来的积极影响更加乐观，但人工智能的伦理风 险并非虚构，普通用户在享受各种创新便利的同时，难免对个人数据滥用、算法决策的不透明心存疑虑，开 发者也担心伦理缺位会使其为由此带来的风险付出高昂代价。为了消除这种双重焦虑，应该通过技术伦理评 估、“技术- 伦理”矫正和信任机制的构建，展开必要的价值伦理校准。更重要的是，要在充分考量人工智能 的社会影响、对区域与全球的兼容和维护和平底线的基础上，构建一种稳健可行的人工智能伦理与治理框架， 实现敏捷治理。

英文摘要:

      While many countries proposing projects and strategies for the development of artificial intelligence（AI），the ethical risks and governance of AI have also become the focus of the common world. Given the different social environments and cultural backgrounds in different regions，it is difficult to build a unified global framework for AI ethics and governance. China is generally more optimistic about the positive impact of AI on the well-being of the economy，society，enterprises，and individuals. However，the ethical risks of AI are not fictitious. On the one hand，while ordinary users enjoy various innovations and conveniences，they will inevitably be afraid of the abuse of personal data and the opacity of algorithmic decisions. On the other hand，developers also worry about the lack of ethical regulation which may make them pay a high price for the risks there arise. To eliminate this double anxiety，it is necessary to carry out“technical-ethical”assessment and correction，and to construct trust mechanism for AI. What’s more，it is important to build a robust and practicable framework for AI ethics，and to achieve agile governance based on full consideration of the social impact of AI，its compatibility within the whole world，and the baseline of maintaining peace.