Chinese electric vehicle (EV) manufacturers are adjusting promotional terminology for advanced driving systems, replacing terms like "intelligent driving" with descriptions such as "assisted driving" to align with stricter regulatory definitions.
Xiaomi recently revised the SU7's driving technology terms from "Intelligent Driving" to "Assisted Driving" in Chinese on its official ordering platforms, which trended on Chinese social media Weibo with more than 22 million views on Monday. In the English version, the standard model retains "Xiaomi Pilot Pro," while the Max version shifts from "Pilot Max" to "Xiaomi HAD" (Highway Assisted Driving).
Under updated guidelines from the Ministry of Industry and Information Technology (MIIT), several major Chinese automakers have adjusted their technical labeling since April. At the Shanghai International Automobile Industry Exhibition, XPeng rebranded its driver-assistance technology as "AI-assisted driving" and announced plans to launch the industry's first "AI Driving Safety Boot Camp," focusing on clarifying system limitations and demonstrating new features through enhanced user training, the Economic Observer reported.
Horizon Robotics named its newly released Level 2 (L2) driver-assistance system "Urban Driver-Assistance System. At Huawei's Qiankun booth, all descriptions of technologies such as "smart driving" and "automatic parking" were revised to "Intelligent Driver Assistance" and "Intelligent Parking Assistance," the China Media Group (CMG) reported.
An industry analyst attributed this terminology recalibration to growing regulatory scrutiny. "Overly optimistic terms like 'fully autonomous' created dangerous gaps between consumer expectations and technological realities," Wu Shuocheng, a veteran automobile industry analyst, told the Global Times on Monday. "Standardized labeling helps reset public understanding."
"These changes reflect China's balanced approach to fostering innovation while prioritizing safety," Wu said. "By tempering marketing hype, regulators create space for sustainable technological advancement."
Zhang Jinhua, president of the China Society of Automotive Engineers, said that most commercially available vehicles in consumer use today are at the L1 or L2 levels. It's important to remember that the L0 to L2 levels represent driver assistance systems, as human drivers remain primary, with machines only playing a supporting role, Zhang told CMG.
As public discussion about smart driving safety intensifies, Wu called for people to have realistic expectations: "While new technologies bring exciting possibilities, drivers should never become complacent. These systems require active human oversight at all times."
Safety is paramount, including the protection of personal and information security, according to Wu. He said that these changes show a significant shift in awareness but stressed that to be truly effective, these guidelines must eventually be formalized into laws and regulations.
On April 16, a department under the MIIT held a meeting to advance the regulation of market access and over-the-air software updates for intelligent connected vehicles. Nearly 60 representatives and major automakers attended.
The meeting emphasized that vehicle makers must rigorously test combined driver-assistance systems, clearly define system boundaries and safety response protocols and avoid misleading or exaggerated claims. They are also required to fulfill disclosure obligations, ensure production consistency and take full responsibility for product quality and safety, with the goal of enhancing the safety of intelligent connected vehicles.
On April 14, the Road Safety Research Institute under the Ministry of Public Security reaffirmed that all commercially available driver-assistance systems in China operate at L2 under national standards, declaring "assisted ≠ autonomous" amid overstatements about smart driving capabilities. The institute urged drivers to strictly respect technical boundaries while demanding that automakers enhance user education and eliminate misleading terminology.
