Browse
Log in / Register
Share
September 15, 2026 – China’s new energy vehicle (NEV) market has crossed a critical threshold: penetration rate above 52%, and L2 driver assistance in over 70% of new passenger cars. Yet at the 2026 International Forum on Chinese Automotive Industry Development (TEDA), the dominant question was not about reaching the next volume milestone. Instead, policymakers, industry leaders and retail analysts debated what will sustain the sector’s upward momentum when scale leadership alone no longer guarantees an edge.
The answer, emerging from multiple panels, centers on a triple shift: from component innovation to system-level AI, from feature stacking to intelligent agent architectures, and from hardware dominance to software-defined mobility solutions.
Opening the forum, Gong Jinfeng, deputy party secretary and general manager of CATARC, delivered a striking data snapshot. In the first eight months of 2026, China sold over 10.65 million NEVs – a penetration rate of 52.4%. Exports hit 6.41 million units through July, up 54% year-on-year. L2 driver assistance now reaches 70.5% of new passenger cars, and the first batch of L3-conditionally autonomous models is running in designated zones.
“China remains the world’s largest single NEV market, with domestic sales accounting for over half of global passenger vehicle sales,” said Gong. He underscored that behind the volume is a broad-based technology acceleration: semi-solid battery energy density has surpassed 400 Wh/kg, solid-state battery R&D has achieved phased milestones, localization of high-compute automotive chips and intelligent operating systems is accelerating, and megawatt-level ultra-fast charging as well as vehicle-to-grid technologies are rolling out faster than expected.
Wang Fang, chief scientist at CATARC, zoomed in on R&D trends. She noted that global NEV sales reached 20.7 million units in 2025, with China selling 13.36 million – its 11th consecutive year at No.1. The rapid L2 penetration jump from 25% (2021) to over 70% now is just the beginning. She identified three battery keywords: safety, charging speed, and solid-state. On the intelligent connectivity side, she sees four vectors: driving, cockpit, chassis, and vehicle-road-cloud collaboration.
Exclusive perspective: Wang also revealed that China’s next-generation battery R&D roadmap explicitly targets commercial solid-state cells by 2028, with several OEMs already validating pilot lines – a timetable that could disrupt global supply chains.
Tian Jietang, director of the Industrial Economy Research Department at the State Council’s Development Research Center, provided a macroeconomic lens. China’s manufacturing share of global output now stands at 27%. With AI taking a global lead, “AI + manufacturing” is creating a unique comparative advantage that no other nation can replicate at scale.
“Manufacturing and innovation are tightly linked – more than 60% of global R&D spending occurs inside the manufacturing sector,” Tian said. He noted that AI’s empowerment is moving from isolated applications like visual inspection and parameter tuning to deep, full-chain integration – from design simulation to supply chain optimization to predictive maintenance.
On the auto industry specifically, Tian made a provocative forecast: “In the next five years, whoever solves motion sickness in the vehicle will own a massive selling point.” He argued that as cars evolve into a ‘third space’ beyond home and office – and potentially into mobile stores – the user experience inside the cabin will become the deciding factor. “Future cars won’t just sell transportation; they will sell scenario-based solutions.”
The diminishing marginal returns of isolated feature stacking are pushing OEMs to rebuild competitiveness from the system level up. Chen Xiaofeng, vice president and chief scientist at Great Wall Motor’s Tech Center, declared that intelligence has entered the era of the “vehicle intelligent agent.”
“Traditional intelligence meant stacking isolated, passive features. Now, AI-driven systems become holistic agents that decompose capabilities in power, chassis, cockpit and driving into atomic services,” Chen explained. Great Wall’s Coffee AI OS 4 architecture combines “large models + AI subsystems + multi-agents + AI interaction.” It achieves three leaps: a “1+N” central brain collaborating with domain experts; open orchestration of more than 4,000 MCP atomic capabilities; and a shift from “users finding services” to “services finding users.”
Zhang Xiaoyu, executive vice president of Changan Automobile, echoed the theme. He defined intelligence as “a vehicle’s ability to self-evolve while holding core technology in-house.” Changan’s Tianshu large model integrates perception, audio-visual processing, and motion control. It upgrades traditional “Clip” training to “Episode” training – a full closed loop from seeing the road and understanding intent to execution and model feedback – creating a continuous evolution logic.
Exclusive detail: Zhang revealed that Changan and BAIC together secured China’s first batch of L3 autonomous driving product licenses. As of early September 2026, Changan users have logged 1.539 billion kilometers, with AEB interventions triggered 15.2 million times and automated parking 12.43 million times.
Wang Chao, vice president of Huawei Digital Energy, shifted the lens to the motion domain – long overlooked in the intelligence race. He argued that innovation in motion control follows an “AI + embodied intelligence” paradigm, where vehicles not only perceive the environment but also physically adapt in real-time. Huawei is investing heavily in distributed drive systems and algorithm-optimized chassis controls that can counter motion sickness while improving energy efficiency.
Gong Jinfeng acknowledged that achievements come with challenges: deepening global trade barriers, regulatory games, and standards battles. He proposed four strategies: target core tech to maintain the lead, deepen cross-sector integration for ecosystem growth, hold the line on product safety, and stick to open cooperation to boost global competitiveness.
Tian Jietang added that China’s advantage lies in its ability to combine manufacturing, AI, scale and policy speed. “The next five years will determine whether China’s auto industry can export not just vehicles, but the architecture of future mobility.”
As the forum closed, a clear consensus emerged: the ascent will not be linear. It will require rewriting the software-defined vehicle stack, rethinking the passenger experience, and integrating battery chemistry with manufacturing AI. For global competitors watching from Munich, Detroit or Tokyo, the TEDA Forum 2026 sent a clear signal – China’s auto industry is leaving the volume game behind and entering a new phase of systemic innovation.
