THURSDAY, MAY 21, 2026 · BRISBANESUBSCRIBE →

THE AI POST

INTELLIGENCE. CURATED.

Robot assembly line in a modern factory
PolicyApril 1, 2026

China Just Wrote the Safety Rulebook for Humanoid Robots. The Rest of the World Does Not Have One.

China has introduced new humanoid robot standards covering safety, autonomy, and system design. The US and EU have nothing comparable.

The AI Post

The AI Post newsroom — delivering AI news at the speed of intelligence.

While Washington debates whether AI chatbots need guardrails, Beijing just published comprehensive safety standards for machines that can walk into your workplace and physically interact with you. The priorities could not be more different.

China has introduced new humanoid robot standards covering safety protocols, autonomy levels, and system architecture. These are not aspirational guidelines. They are technical standards designed to govern an industry that is already producing thousands of units per month on Chinese factory floors.

The timing is not accidental. Agibot just hit 10,000 humanoid robots produced. UBTech plans 5,000 units this year and 10,000 next year. A McDonald's in Shanghai is testing robot servers. The humanoid robot industry in China is no longer a research exercise. It is a mass-production reality that needs rules before something goes badly wrong.

The Safety Case Is Not Theoretical

Industrial robot incidents already happen in controlled environments with safety barriers and strict procedures. In 2023, a Tesla Fremont technician was injured when a robot arm activated unexpectedly during maintenance. Similar incidents have been reported across the automotive sector for years.

Humanoid robots change the equation entirely. They are designed to share physical space with humans, navigate unpredictable environments, and make autonomous decisions about movement and interaction. A traditional industrial robot bolted to the floor behind a cage is a known risk. A humanoid robot walking through a restaurant or warehouse is a category of risk that existing safety frameworks were never built to address.

The Regulatory Gap Is Enormous

The United States has no federal framework for humanoid robot safety. The EU AI Act addresses high-risk AI systems but was not designed with autonomous bipedal machines in mind. Japan and South Korea have advanced robotics industries but no published standards specifically for humanoids in public spaces.

China, by moving first, gains two advantages. First, its domestic manufacturers will design to these standards from day one, building safety into the product rather than retrofitting it later. Second, in the absence of competing international standards, Chinese specifications could become the de facto global baseline, particularly for countries importing Chinese-made humanoid robots at rapidly declining prices.

This is the regulatory playbook China used with 5G infrastructure and electric vehicles. Write the standards. Scale manufacturing. Export the ecosystem. By the time other countries catch up on regulation, the supply chains are already locked in.

The West spent the last two years arguing about whether AI can write poetry that infringes copyright. China spent it building safety standards for robots that can crush a human skull. One of these problems is more urgent than the other.

Sources: Robotics and Automation News, Agibot, UBTech.

Chinahumanoid robotsrobot safetyregulationstandardspolicy