扩展的东盟人工智能治理和伦理指南——生成型人工智能（英）

Investment Rating - The report does not explicitly provide an investment rating for the industry Core Insights - The Expanded ASEAN Guide on AI Governance and Ethics focuses on generative AI (Gen AI) and outlines the opportunities and risks associated with its adoption in the ASEAN region, emphasizing the need for responsible practices and policy actions to ensure safety and promote benefits [2][19] Summary by Sections Executive Summary - The document serves as a supplement to the ASEAN Guide on AI Governance and Ethics (2024), providing policy considerations for Gen AI and recommending actions for its responsible adoption [2][4] Gen AI Risks - The report identifies six key risks associated with Gen AI, including mistakes and anthropomorphism, factually inaccurate responses, deepfakes, infringement of intellectual property rights, privacy issues, and the propagation of embedded biases [5][19] Policy Recommendations - A range of policy recommendations is defined for ASEAN to promote trusted and responsible use of Gen AI, emphasizing a coordinated, pro-innovation regional response [4][21] Use Cases - The guide illustrates the implementation of its policy recommendations through four detailed use cases involving public and private institutions in ASEAN that align with AI governance and ethics [7][8] Guiding Principles - The report outlines guiding principles for fostering trust in AI, which include transparency, fairness, security, human-centricity, and privacy, all of which are crucial for the ethical design and deployment of Gen AI systems [26][39][43] Accountability and Integrity - Emphasizes the need for human accountability in the design, development, and deployment of Gen AI systems, highlighting the importance of governance mechanisms for oversight [49][50] Robustness and Reliability - Stresses that Gen AI systems should be robust enough to handle errors and unexpected inputs, and that developers should create measurable benchmarks for acceptable outcomes [52][53]