Beijing, October 9, 2026. China issued new government guidelines on Friday calling for AI that is “safe, reliable and controllable” — and warning that officials who waste money on “blind” AI investment will be held strictly accountable.
What the guidelines say
According to state-run Xinhua, which reported the new guidelines on Friday, China will build technology monitoring, risk warning, and emergency response systems designed to keep AI safe, reliable, and controllable. The guidelines frame AI as one of the country’s “new productive forces” — a group of frontier technologies Beijing wants to accelerate without losing oversight.
The sharper edge is financial. Authorities will “hold strictly accountable” those responsible for major losses caused by “blind” investment in these technologies. In plain terms: Beijing has watched speculative AI spending and is signaling that local officials and state-backed actors who throw money at hype could face consequences.
The low-altitude economy gets a nod
The guidelines also call for the “healthy and orderly” development of China’s low-altitude economy — the businesses built around civilian drones, air taxis, small aircraft, and related services operating in low-altitude airspace — while ensuring safety. It’s a sign that Beijing sees AI’s physical applications (drones, autonomous flight) as part of the same governance push.
Why it matters
Timing is everything here. These guidelines landed on the same day a SemiAnalysis report found that just 3.6% of China’s leading AI models have publicly published safety tests — with no dangerous-capability testing (cyber, bio, loss-of-control) publicly disclosed by any major Chinese developer. Beijing is now officially demanding safety systems it doesn’t yet have evidence of.
For Western readers, this is the other half of the AI race: the U.S. and Europe argue about voluntary guardrails and bills with teeth, while China governs AI through top-down accountability — monitoring systems plus personal responsibility for wasteful spending. Both sides are still building the plane while flying it; they just have different checklists.
FAQ
What are “new productive forces”?
It’s Beijing’s official term for the frontier technologies it wants driving the next wave of economic growth — AI chief among them, alongside other advanced tech.
What does “strictly accountable” mean in practice?
The guidelines don’t specify penalties in Xinhua’s summary. In China’s governance system, though, that phrasing typically signals that officials and state-sector managers could face career or legal consequences for major wasteful investments.
Does China already test its AI models for safety?
Barely, according to independent research: a SemiAnalysis review of 857 models from nine leading Chinese AI companies found only 31 had published model-specific safety evaluations — and just nine at or before launch.
What’s the low-altitude economy?
Businesses operating in airspace below roughly 1,000 meters — delivery drones, air taxis, small aircraft and the services around them. China is treating it as a strategic growth sector tied to its AI push.
Sources: Reuters, Xinhua.