According to a news release from the Cyberspace Administration of China, the Cyberspace Administration of China and the National Development and Reform Commission have recently jointly issued the "Guidelines for Deploying and Applying Large-Scale AI Models in Government". These guidelines serve as a reference framework for government departments at all levels when deploying and applying large-scale artificial intelligence (AI) models. The "Guidelines" underscore the importance of standardized deployment practices. They stipulate that government entities should thoughtfully select implementation strategies for large-scale AI models, taking into account the unique demands of various governmental contexts and the existing technological infrastructure.
Deployment efforts ought to embrace a coordinated and centralized strategy. Specifically, cities are mandated to align with unified provincial directives, while counties and lower administrative levels should, as a general rule, leverage superior intelligent computing capabilities and model resources already in place. Concurrently, there should be a proactive exploration into establishing a centralized deployment paradigm characterized by "one-time construction for shared use across multiple locations and departments". This approach aims to avert the emergence of isolated "model silos". Additionally, the guidelines advocate for bolstering government data governance initiatives to elevate data quality, thereby providing robust support for the refinement and training of large-scale governmental AI models.