Machines

The identity cards of Chinese robots raise questions for the EU

In a sector lacking leading brands, traceability becomes a competitive advantage

by Marco Gervasi

Reuters

3' min read

Translated by AI
Versione italiana

Key points

  • Beijing's move

3' min read

Translated by AI
Versione italiana

China has just launched a new system to assign a 29-character code to over 28,000 humanoid robots, which will track them from the factory right through to recycling. This code records the manufacturer, model, hardware specifications, level of artificial intelligence capability, installed software, component wear and tear, and the entire operational history. It is not a static record, but a living digital profile: it tracks joint wear, battery degradation and movement precision in real time, and when something goes wrong, it allows the fault to be quickly identified.

Beijing's move

It is a tracking infrastructure that does not exist in the West. The European AI Act classifies artificial intelligence systems according to risk, but does not provide for the identification of individual physical robots. Not even the United States has anything like it. Put simply, China has just built what we are still lacking.

Loading...

It is no coincidence that Beijing is leading the way. The country is home to more than a hundred humanoid robot manufacturers, and in 2025 it invested $3.4 billion in new start-ups in the sector – 42% more than the United States and five times the European total. Robots have already left the laboratories: a humanoid robot has independently completed a half-marathon in Beijing, the electricity giant State Grid plans to deploy thousands of them for plant maintenance, and the first trials with robotic workers have begun in the tea plantations of Hubei. When robots start to number in the tens of thousands, the question of who each machine is ceases to be theoretical.

The issue of machine liability

What is driving China is, in fact, a problem that will soon be ours too: liability. When a robot injures a worker or causes damage, it must be possible to trace the machine responsible and identify who built it. This is precisely the issue that Europe will tackle from 2027 with the new Machinery Directive: Beijing has resolved it on a technical level, whilst here it remains subject to liability rules. The Chinese system, moreover, does not stand alone: it continues a line of regulation that, since 2022, has already governed recommendation algorithms, generative artificial intelligence and synthetic content, now extending the same lifecycle monitoring logic to physical robots. These are, in fact, treated in the same way as cars or medical devices.

This raises an issue of particular interest to those looking to invest. In a Chinese market with over a hundred manufacturers and no dominant brand, a robot with a verifiable track record becomes a sign of quality for buyers. A well-maintained unit, regularly updated and used within the limits for which it was designed, carries with it a history that attests to its reliability. Compliance, which began as an obligation, thus becomes a competitive advantage: in a sector still lacking leading brands, traceability becomes a tool for competition even before it is a regulatory requirement. This means that Chinese robots could arrive in Europe already ‘traceable’, with a verifiable history that competitors are unable to offer.

The same approach can be seen, at a local level, in the city of Hangzhou, where, from May 2026, a law specifically governing intelligent robots will introduce code-based traceability and test environments with flexible regulations. These are tools that the AI Act provides for in theory, but which are already operational in China, centred around an industrial district that in 2025 reached a production value of $15.7 billion and is home to over 700 companies. Observers regard it as a possible model for the country’s future legislation.

A comparison with Europe

However, there remains a gap in values that should not be overlooked. The Chinese ID system — explicitly modelled on the national register of citizens — demonstrates how the state has oversight of vehicles and technology in a way that would be problematic in Europe from the perspective of privacy and the GDPR. Our interest, therefore, is not in ‘imitating’ the Chinese model, but in understanding what this system actually achieves — accountability, traceability, trust — and what, on the other hand, conflicts with the principles on which our market is founded. The robot, incorporating AI, will have to adhere to the rules of the territory in which it is deployed, but it will certainly be able to adopt the technical innovations that improve its safety and transparency. The real question, then, is not whether the Chinese approach is right, but whether it is simply ahead of its time. Should humanoids become as commonplace as China anticipates, every country will need a way to recognise them.

marco.gervasi@theredsynery.com

Copyright reserved ©
Loading...

Brand connect

Loading...

Newsletter

Notizie e approfondimenti sugli avvenimenti politici, economici e finanziari.

Iscriviti