China’s AI-Automotive Integration Reaches New Heights
Artificial intelligence is accelerating its integration into China’s automotive industry, transforming vehicles from mere transportation tools into intelligent terminals and mobile living spaces. The recent approval of the world’s first global technical regulation for autonomous driving systems—led by China—marks a new phase in the country’s smart vehicle sector, according to Xinhua News.
Technology giants including Huawei, Tencent, and ByteDance are actively partnering with mainstream automakers, deploying AI across the automotive value chain and reshaping the industry’s competitive landscape.
Smart Driving Accelerates
At night on a highway, a vehicle traveling at 120 km/h can identify an obstacle the size of a soda can from 100 meters away, three seconds in advance, and plan an evasive route. This is the capability of the Qijing GT7, which launched recently with the latest version of Huawei’s Qiankun ADS assisted driving system, as reported by Sina Finance.
As intelligent connected vehicle technology iterates at an unprecedented pace, driving is shifting from human-dominated to human-machine collaborative. Huawei’s Qiankun ADS has accumulated over 12.2 billion kilometers of assisted driving mileage, with vehicles equipped with the system logging more than 35 billion kilometers in total.
“Whether users like to use it is the only standard for whether assisted driving is good,” a Huawei Intelligent Automotive Solutions executive said.
Li Mingyan, director of the Intelligent Economy Research Institute at SAI Industrial Research Institute, noted that AI is fundamentally reshaping intelligent driving, with the most significant change being the industry-wide shift toward data-driven end-to-end large model architectures that make vehicle behavior more human-like and better equipped to handle complex edge cases.
The industry is also converging on a “heavy perception, light map” approach, reducing reliance on frequently updated high-precision maps in favor of enhanced real-time perception capabilities. VLA (Vision-Language-Action) large models are emerging as a frontier direction, unifying vision, language, and action into a single framework that enables systems to understand semantic information such as traffic police gestures and temporary road signs.
User Experience Transforms
From transportation to mobile living space, smart vehicles are transforming the user experience. Voice-controlled navigation, entertainment, and vehicle functions have become standard in-car interactions.
“The cockpit is moving toward the Agentic AI era,” said Zhong Xuedan, Tencent’s vice president of smart mobility. “AI is no longer about stacking features, but organizing into a perceptible, planable, executable, and continuously evolving intelligent hub.”
Tencent holds a 49% market share in “City NOA smart driving maps” for new energy passenger vehicles, according to Gaogong Intelligent Vehicle Research Institute data. Tencent Cloud’s computing solutions for automakers achieve 98.4% utilization of hybrid computing clusters and over 60% GPU efficiency improvement in inference scenarios.
ByteDance’s Volcano Engine has partnered with 100% of mainstream automakers. The Doubao large model is installed in over 7 million smart vehicles across more than 50 car brands and 145 models—the highest in the industry—completing over 30 million cockpit interactions and service loops daily.
Zhang Linshan, a researcher at the NDRC Macroeconomic Research Institute, said AI large models have brought comprehensive revolutionary changes to intelligent cockpits and human-vehicle interaction, with intelligence shifting from an “optional feature” to a core consideration in car purchases.
Safety Standards Tighten
China has also moved decisively to establish safety guardrails. On July 30, the Ministry of Industry and Information Technology released the mandatory national standard “Safety Requirements for Intelligent Connected Vehicles - Autonomous Driving Systems” (GB 44721-2026), which takes effect on July 1, 2027, as detailed by OFweek.
This milestone follows China’s leadership in developing the world’s first global technical regulation for autonomous driving systems (ADS GTR), approved in June at the UN/WP.29 199th plenary session in Geneva. Jointly led by China, the EU, UK, US, Canada, and Japan, the regulation provides unified standards for safe deployment of autonomous driving technology worldwide, as reported by Interface News.
“How to balance intelligent innovation with safety and reliability is a systems engineering and governance proposition,” Li Mingyan said, recommending a design philosophy of safety-first, progressive innovation where new features are built on thorough risk assessment, simulation testing, and closed-field verification.
Industry Transformation
The AI-automotive convergence is driving a fundamental industry shift. Core value is moving from traditional mechanical hardware—engines and transmissions—to software, algorithms, and data processing capabilities. Software-defined vehicles have become a reality.
L2-level high-end intelligent driving has achieved large-scale commercial deployment, with penetration of combined driving assistance systems in new passenger cars exceeding 60%. L2 and L2+ penetration rose from 60% in 2024 to 88% in 2025, with projections to exceed 95% by 2026-2027, according to OFweek analysis.
China conditionally approved its first two L3-level vehicle models in December 2025, and L3 autonomous driving is steadily advancing through pilot access programs.
Zhang Linshan emphasized that AI is restructuring R&D, production, and supply chain processes—shortening vehicle development cycles, optimizing quality control, and shifting the industry from one-time vehicle sales to long-term service monetization, transforming traditional manufacturing into “manufacturing plus services.”
What’s Next
As the industry moves toward 2027, when GB 44721-2026 takes effect, the focus will be on scaling high-level autonomous driving, conquering extreme edge-case scenarios, and establishing sustainable commercial models. The convergence of AI with automotive manufacturing positions China at the forefront of a global transformation—one where vehicles are no longer just machines, but intelligent companions on the road.