China’s MIIT unveils 2026–2028 AI plus telecoms roadmap

Beijing’s latest ‘AI + information and communications’ plan shows how seriously China is taking the fusion of AI with its network infrastructure. Rather than treating AI as a layer on top of 5G and cloud, the document envisages intelligent wireless interfaces, space‑based computing networks and an “internet of agents” as core national capabilities. The emphasis on building end‑to‑end platforms—compute, data, models and applications—signals that Beijing wants domestic equivalents of the full‑stack AI‑cloud offerings now being rolled out by U.S. hyperscalers.

For the race to AGI, this is less about a single breakthrough and more about sustained capacity building. If MIIT succeeds in knitting AI tightly into China’s telecoms fabric, Chinese labs and companies will enjoy abundant, low‑latency access to edge compute, data and users. That, in turn, can accelerate experimentation with agentic systems, embodied AI and real‑time multimodal services, all of which are likely stepping stones toward more general intelligence.

The policy also has a defensive angle: by insisting on domestic AI model platforms and ‘sovereign’ compute, China is hedging against export controls and supply‑chain shocks. From a competitive standpoint, this raises the bar for any foreign AI or cloud provider hoping to play in China’s telecom and industrial markets—they will be expected to slot into, not define, the national AI‑ICT architecture.