“构建人工智能护栏的全球共识”课题项目启动

Group 1 - The project "Building a Global Consensus on AI Safeguards" aims to establish a clear, actionable, and sustainable consensus mechanism for AI governance, addressing divergences in AI governance frameworks across countries [2] - The project will utilize a "panoramic mapping" approach to identify consensus, divergences, and gaps in AI governance, and will implement a dynamic tracking mechanism for identifying global AI governance safety lines and common foundations [2] - The project emphasizes a full-process mechanism of "dialogue—co-research—transformation—tracking" to promote the consolidation of AI governance consensus and sustainable technological development [2] Group 2 - Experts highlighted the need to differentiate between "existential risks" and "systemic risks" in AI, advocating for current policies to focus on solvable systemic risks rather than hypothetical long-term threats [2][3] - The importance of establishing a "PDCA" (Plan-Do-Check-Act) closed-loop management mechanism from enterprises to nations was emphasized to address uncertainties arising from rapid technological iterations [3] - The necessity for international cooperation in identifying non-controversial governance topics was suggested to overcome current dialogue challenges in AI governance, which are hindered by discourse divergences [3]