China MIIT Sets 75% Metro AI Coverage Target for 2028
Key insights
- China's MIIT set a 75% metropolitan coverage target for one-millisecond-latency computing access by 2028.
- The three-year plan covers 17 tasks across four areas including intelligent industry upgrading, integrated applications promotion, and industry governance.
- A 2030 second horizon targets major advances in core AI-communications technologies and a complete collaborative innovation ecosystem.
Why this matters
China's MIIT is treating its national telecoms network as the primary AI delivery substrate, anchoring the plan to hard latency and coverage metrics rather than aspirational language. The 75% metropolitan coverage target for one-millisecond computing access creates a concrete 2028 deadline for network and computing vendors operating in China to demonstrate compliance. For AI practitioners building latency-sensitive applications, this plan maps the infrastructure horizon: widespread sub-1ms computing access across Chinese metros by 2028, with core technology advances targeted for 2030.
Summary
China's MIIT released a three-year plan June 10 to embed AI into the country's information and communications sector through 2028, with hard numeric targets as the accountability mechanism.
The plan covers 17 tasks in four areas: intelligent industry upgrading, AI foundations, integrated applications, and governance. Key metrics: 75% metro coverage at one-millisecond latency and over 30 specialized agents and high-value use cases by 2028.
Essentially: China's MIIT is wiring AI performance requirements into the telecom network layer, not just data centers.
- 75% of metropolitan areas must have sub-1ms computing access by 2028
- 30+ specialized intelligent agents and high-value use cases across the ICT sector
- A 2030 horizon targets core technology breakthroughs and a collaborative innovation ecosystem
The latency and coverage targets make this a delivery framework with measurable checkpoints, not a policy statement.
Potential risks and opportunities
Risks
- If 75% metro coverage slips past 2028, AI applications built around sub-1ms latency assumptions face delayed deployment across China's ICT sector with no stated fallback timeline.
- The gap between the 2028 network readiness target and the 2030 core technology milestone creates a window where infrastructure may be declared AI-ready before the underlying AI capabilities are mature.
- Telecoms vendors integrating with Chinese networks face a moving compliance target as the 17-task plan generates technical standards without specifying which bodies validate compliance or by when.
Opportunities
- Edge computing and network infrastructure vendors able to demonstrate sub-1ms computing delivery gain a government-mandated procurement argument through 2028.
- AI model optimization vendors targeting the large-and-small model collaboration technologies named in the plan gain a defined procurement window with MIIT as an implied buyer.
- AI use-case integrators across the ICT sector benefit from a 30+ application mandate that defines concrete deployment scope through 2028.
What we don't know yet
- No budget or investment figure is specified for implementing the 17 tasks, leaving the scale of funding unclear for the 2026-2028 period.
- Which specific sectors or industries fall within the 30+ mandated use cases is not detailed in the plan.
- How the 75% metropolitan coverage threshold will be measured and enforced across varied regional infrastructure is not addressed.
Originally reported by english.news.cn
Read the original article →Original headline: China's Ministry of Industry and Information Technology Releases 3-Year AI-ICT Integration Plan — 75% Metro Coverage at 1ms Latency, 30+ AI Use Cases by 2028