Regulatory sandboxes for autonomous driving in China: Experimental governance, open innovation, and legal institutionalization

The rapid development and commercialization of autonomous driving technology pose systemic challenges to traditional legal and regulatory frameworks. Regulatory sandboxes have therefore emerged as important governance instruments for addressing technological uncertainty while balancing innovation with safety. This article argues that the autonomous driving regulatory sandbox should not be confined to a mere “safe testing space” but should be understood as an open-innovation governance infrastructure that integrates experimental regulation, data-driven learning, and multi-stakeholder coordination. Through doctrinal analysis and functional comparison, the article demonstrates that the autonomous driving regulatory sandbox has three core institutional attributes: an experimental regulatory zone, a legislative data pool and a multi-stakeholder co-governance platform. It further examines the limitations of China’s current access-based pilot programs, particularly their weaknesses in regulatory experimentation, legislative learning and collaborative governance. These limitations reveal deeper legal barriers concerning legislative authorization, fragmented regulatory structures, liability allocation and data governance. On this basis, the article proposes an institutional framework for China’s autonomous driving regulatory sandbox, covering legislative empowerment, governance structure, liability and risk allocation, and data governance mechanisms. The article contributes by showing how regulatory sandboxes can support open innovation, institutional learning and legally accountable experimentation in emerging technology governance.