_________

China's car industry has mostly accomplished its transformation to electric and intelligent vehicles. Now, the focus turns to a third wave of transformation with artificial intelligence.

The Ministry of Industry and Information Technology (MIIT) in Beijing has just published a new five-year plan which highlights “AI + Automobiles” as a priority for the first time.

Among the categories it emphasizes are AI agents, integration of cockpit and driving functions, and shared industry data. In general, it asks companies to “accelerate” the deployment of AI across the whole value chain of the automotive industry.

Industry insiders caution that this AI wave is in its early days, however. A recently published white paper states that while more than 60 automakers have by now deployed DeepSeek locally, that alone cannot be called a transformation to the AI-driven vehicle. The authors mockingly call that “technology rebranding”.

The transformation has only just begun, but Chinese carmakers are clearly embracing it as their next chance to gain a competitive advantage globally. As cheap batteries, electric drivetrains and even intelligent driving features become standard, the competitive edge is shifting toward creative use of AI.

And exactly because it is still early days in this third big disruption of carmaking in China, European and other foreign companies are looking at tremendous business opportunities. Siemens, for example, offers more than 100 AI agent applications for the Chinese market.

In this week's Must Read, we explore Chinese and foreign companies' efforts in three key areas of the AI transformation, their challenges and their emerging strengths.

_________

Alibaba used its annual Aspara Conference in Hangzhou to lay out the most ambitious infrastructure plan any Chinese company has announced so far. On September 22, 2026, its chip subsidiary T-Head unveiled the Zhenwu V900, an AI accelerator with three times the computing performance of its predecessor, the Zhenwu M890. A single cluster can scale to as many as 500,000 chips.

The chip is one of three pillars. “If tokens are the electricity of the AI era, chips are the generators,” Alibaba CEO Eddie Wu said.

With regard to the cloud, Alibaba aims for more than 20 gigawatts of global data center capacity by 2032, starting with a new site in the Netherlands in October.

Alibaba is assembling a domestic full stack of models, chips and cloud at the moment when Washington is trying to keep advanced silicon out of China.

State-owned enterprises (SOEs) in China are getting serious about robots. They commission them from private robot makers and also invest real money in their own robot manufacturing.

State Grid has an ice-removal robot that a drone lifts onto a high-voltage line. Another SOE, CSSC Chongqing Equipment, built a humanoid that sprays insulation inside ship compartments, a job for which workers are hard to find.

In August, Norinco Group formally established a robot innovation consortium of central state-owned enterprises with more than 100 organizations..

The entry of the state giants answers a question that hung like a shadow over China’s robot boom, namely who will buy the machines.

Envision, a Chinese green energy company and wind turbine maker, is connecting AI data centers directly with green wind power. The biggest project of this kind so far has now fired up its servers and promises to deliver exceptionally cheap tokens.

Galaxy Campus can run one million chips in parallel and has computing power on the scale of one million petaflops, which makes it the single AI data center with the highest token output capacity in the world, according to Envision.
Envision’s founder Zhang Lei wants to make tokens “as cheap as potatoes.”

Cheap tokens are the basis of China’s strategy to make capable AI ubiquitous.
This week’s Deep Dive looks at the European suppliers, Siemens and ABB among them, that sell the power technology such campuses need.

Alibaba’s DAMO Academy has released DAMO EAGLE, an AI model that screens for esophageal cancer using nothing more than a routine chest CT scan.

The model was developed with Sichuan Cancer Hospital and Sun Yat-sen University Cancer Center and needs neither intubation nor contrast agents.

EAGLE was trained on 6,813 patients and validated across 12 centers in three countries with 80,612 patients. It is Alibaba’s fourth cancer-screening model, after pancreas, stomach and colorectal cancer.

China accounts for roughly half of the world’s esophageal cancer cases and deaths, and most patients are diagnosed late. Even though false positive-rates and other problems need to be monitored, reducing the cost of cancer screening is benefitial.

For details and more analysis, please click here to read the full, free version of this week’s Highlights:

_________

_________

AI applications now cover 34.2 percent of key manufacturing scenarios in China.

Source: Xinhua News Agency

China had 4.95 million active AI agents in 2026. Projections suggest this could soar to 197 million by 2030, according to the 2026 China Artificial Intelligence Computing Power Development Assessment Report. This implies a compound annual growth rate of 151.2 percent.

Sources: Securities Times (Zhengquan Shibao)

Alibaba aims to boost its global data center capacity, managed by Alibaba Cloud, to over 20 gigawatts by 2032. CEO Eddie Wu announced this at Alibaba’s Apsara Conference on September 22, 2026.

Source: Huaerjie Jianwen

_________

Shenzhen’s courts have made artificial intelligence a routine tool for their judges. Large AI models and agents have supported more than 700,000 cases since the system was launched in June 2024, reported the newspaper Nanfang Dushi Bao on September 25, 2026.

The figures come from a white paper released at the China International Forum on Rule of Law, held in Shenzhen on September 22 and 23. The system covers civil, criminal and administrative litigation. By August 31, 2026, it had helped review the files of 609,000 cases, supported 269,000 court hearings and assisted in drafting 678,000 legal documents.

Judges closed an average of 744 cases each in 2025, which was 249 more than in the year before. The white paper also claims gains in quality. The appeal rate fell by 48 percent. The rate of judgments overturned or sent back for retrial fell by 65 percent. The report does not name the period of comparison for these two figures.

The white paper describes the system as the world’s first systematic AI-assisted judicial system. “Shenzhen is one of the earliest places in the world where a court system has systematically applied large artificial intelligence language models in real-world working environments to assist with judicial work,” said Liu Zhuang, associate professor at the University of Hong Kong Faculty of Law.

Judicial and government delegations from 20 countries, including Brazil, Vietnam, Singapore and Portugal, have visited Shenzhen to study its practices. The city issued an AI action plan for 2026 to 2028 on September 4. It names governance as one of six areas in which AI is to be deeply integrated by 2028. The core AI industry is to surpass 300 billion yuan (about US$44 billion) by then.

Foreign companies that sue or are sued in Shenzhen will meet judges who work with AI. The system reviews files, supports hearings and helps draft legal documents. Legal departments can look forward to a faster pace in court.

China’s Ministry of Transport has issued safety guidelines for AI applications in its sector. The Guidelines for the Safe Application of “Artificial Intelligence + Transportation” were approved, issued and implemented on August 27, 2026, according to an explainer published on the ministry’s website on September 14, 2026 .

The text has the status of a national standardization guidance document. It is meant to improve safety rules for AI applications in transportation and to address the risks arising from them. AI applications in the sector are to be safe, reliable and controllable, the ministry wrote.

The guidelines belong to a wider initiative. The ministry has issued development guidelines for a first batch of 41 typical application scenarios in 10 areas. Among them are intelligent driving, smart highways, intelligent railways, smart civil aviation and smart postal services. Work on the first batch of scenario projects has officially begun.

Transportation is one of the sectors at the forefront of AI adoption in China. Safety guidance now arrives alongside the first 41 application scenarios.
Carmakers, logistics companies and suppliers of rail and aviation technology should read the guidelines before they take part in scenario projects.

China opened its 2026 Cybersecurity Week in Jinan on September 14 with the release of the Artificial Intelligence Security Governance Framework 3.0, reported the state broadcaster CGTN.

The event ran until September 20 and was jointly organized by 10 government departments and organizations. Among them are the Cyberspace Administration of China, the Ministry of Industry and Information Technology and the Ministry of Public Security. This year’s theme was the “safeguarding of cybersecurity in the intelligent era”.

Two further documents were released. One contains the 2026 test results for AI-enabled cybersecurity applications. The other lists consumer-grade networked cameras that have completed cybersecurity labeling registration. CGTN did not report what has changed in version 3.0 of the framework.

The framework has reached its third version, and Beijing chose the opening of its national cybersecurity week to present it. AI security is treated as part of cybersecurity, with the cyberspace regulator, the industry ministry and the police among the organizers.

Companies that run AI systems in China should obtain the full text and compare it with their own safeguards.

Guangdong province has launched a plan to enhance public scientific literacy from 2026 to 2030, reported Guangdong Keji Bao on September 23, 2026.

A key project is the "AI + Science Popularization" initiative. Science and technology professionals will receive extensive AI training. AI will also play a significant role in science communication and education.

Another project focuses on industrial workers. The province aims to develop standards for new occupations and skills training. Workers will enhance their ability to use AI and the industrial internet.

AI in factories relies on skilled workers. Guangdong has now prioritized this in a five-year plan. Foreign manufacturers in the province should monitor new occupational standards, as they will influence training and hiring practices.

_________

Artificial intelligence was high on the agenda when Donald Trump and Xi Jinping met in Washington this week. “Whoever wins AI, WINS!” Trump declared.

Xi Jinping chose more diplomatic words. The two countries “should be partners rather than rivals,” he wrote in a statement upon landing. Questions of style aside, both governments are locked in an AI race that is more and more reminiscent of the Cold War.

Far away from such headlines and all the political noise, a number of successful European companies stay focussed on what they do best. Business.

Siemens software runs the “smart brain” of a new plant in Liuzhou where humanoids assist with making humanoids for UBTech. Schaeffler supplies bearings, gears and motors to the Chinese company Leju Robotics. ABB sells the power technology that gigawatt-scale computing centers need.

The AI race between the U.S. and China is as unfortunate as it is real, but it is not in Europe's interest to pick a side. What China’s AI boom needs next are factories, electricity and precision instruments. Those are exactly the things European industry has been selling for a century. Getting dragged into an antagonistic posture toward Beijing by Washington would cost Europe a market with enormous growth potential.

What the digital twin that Siemens built with UBTech actually does before the first humanoid is assembled, how Zeiss optics act as the eyes of Foxconn’s AI inspection, how ABB makes good money on the 800-volt direct current now feeding data centers in China, and what a 2-gigawatt computing campus in Inner Mongolia powered by wind and solar tells us about the size of the opportunity, all of this is the subject of this week’s Deep Dive.