The black box and the rulebook: China’s quest for proportionate explainability in automated governance

Chinese scientists and legal scholars have begun to address one of the most pressing challenges of the digital age: how to make automated administrative decisions explainable without paralyzing the systems that deliver them. Pengfei Shen’s forthcoming study, published in Computer Law & Security Review, proposes a proportionate explainability framework specifically designed for China’s rapidly evolving governance landscape.

The research acknowledges a fundamental tension. On one hand, China’s administrative agencies increasingly rely on algorithmic systems to process vast numbers of routine decisions—from social credit assessments to licensing approvals—with speed and consistency that human adjudicators cannot match. On the other hand, citizens affected by these decisions have legitimate claims to understand the reasoning behind them, particularly when errors or biases may be embedded in opaque machine learning models.

Shen argues that a one-size-fits-all explainability mandate would be both impractical and counterproductive. Instead, the proposed framework calibrates the required level of explanation according to the decision’s impact, the complexity of the algorithm, and the specific legal context. For low-stakes administrative matters, a simplified justification may suffice; for high-impact determinations affecting property, liberty, or reputation, the standard of explainability must rise proportionately.

This is not merely an academic exercise. China’s Personal Information Protection Law and its emerging regulatory architecture for AI already embed requirements for transparency and the right to explanation. Shen’s work provides the doctrinal scaffolding to operationalize those principles in the administrative context, offering guidance to lawmakers, regulators, and system designers. The proportionate approach could be particularly influential as other jurisdictions—including the European Union under its AI Act—struggle with similar questions about algorithmic accountability in the public sector.

Why it matters:
As governments worldwide deploy AI for administrative decisions, China’s effort to institutionalize proportionate explainability offers a practical template for reconciling algorithmic efficiency with procedural justice. The framework’s reception in China could shape global standards for algorithmic transparency in the public sector.

Source →