Tsinghua University officially inaugurated the Qingzhi AI Security Hub and hosted an International Symposium on AI Safety on July 15 at its Beijing campus. The launch underscores a strategic commitment to advancing artificial intelligence governance and aligns with global efforts to mitigate emerging technological risks. The hub aims to establish a premier, internationally recognized research center focused on AI safety, risk assessment, and technical standardization. Operating from the Qingzhi Park campus, it will prioritize critical research domains including system alignment, model evaluation, interpretability, and the modeling of control risks in autonomous systems. University leadership emphasized that the initiative responds to both national strategic priorities and the broader global technology governance landscape. Core operational pillars will include a shared research workspace, programs for visiting international scholars, structured talent development for early-career researchers, and mechanisms for cross-institutional exchange. Academician Yao Qizhi, Dean of the AI School and the Institute for Interdisciplinary Information Sciences, highlighted pressing challenges such as the potential loss of control during long-horizon autonomous tasks, the cognitive impacts of AI on human decision-making, and emergent cybersecurity and biosecurity threats posed by advanced models. The inauguration was attended by senior university officials, industry representatives, and international experts. Sören Mindermann, Technical Research Lead at the Cambridge University Centre for the Alignment and Safety of AI and Policy, noted that establishing a unified scientific framework is complex but essential. He praised the hub’s residency and collaboration programs as vital infrastructure for sustaining multilateral research efforts. Following the ceremony, the accompanying international symposium convened researchers, policymakers, and industry leaders for plenary presentations and closed-door roundtable discussions. Central themes included the translation of theoretical safety research into industrial deployment, the operationalization of safety protocols at scale, and strategies for harmonizing domestic and international regulatory frameworks. The establishment of the Qingzhi AI Security Hub marks a significant step in Tsinghua’s expansion of cross-border academic cooperation and frontline AI safety research. By integrating academic innovation with industry application, the hub is positioned to contribute substantially to global AI risk governance and multilateral technical collaboration. The initiative reinforces institutional capacity to anticipate, evaluate, and mitigate the systemic risks associated with next-generation artificial intelligence systems.