China outlines ‘AI + human resources’ plan with 2030 adoption goals

China’s new “AI + human resources and social security” roadmap is a classic example of how Beijing turns abstract tech priorities into quantified deployment targets. By 2026, the plan calls for foundational infrastructure and 20 application scenarios built on sector‑specific models; by 2027, dozens more high‑value use cases; and by 2030, pervasive AI use across HR and social security. That’s less about frontier capability and more about building large, labeled datasets and real‑world harnesses in one of the most data‑rich bureaucratic systems on earth.

For AGI watchers, the significance is twofold. First, at this scale, administrative AI systems become a proving ground for long‑horizon, high‑stakes decision‑making with human lives and welfare on the line—exactly where alignment and robustness are hardest. Second, embedding AI into state social systems deepens the feedback loop between governance and model development: the state sets targets, models are trained on citizen data, and outcomes shape future policy. Over time, that could give Chinese labs a unique edge in building models tuned for complex institutional decision‑making, even if they lag slightly on headline benchmarks.