How AI Is Teaching China’s Cars to Read Emotions

Artificial intelligence is transforming China’s automotive industry, enabling vehicles to move beyond passive voice commands and into a new era where cars can see, hear, and proactively anticipate the needs of drivers and passengers. The shift from “software-defined vehicles” to “AI-defined vehicles” was the central theme of the 2026 Qualcomm Automotive Technology and Cooperation Summit, held June 4–5 in Wuxi, Jiangsu Province, where over 70 automotive electronics suppliers showcased more than 50 real-vehicle demonstrations.

The Rise of AI-Defined Vehicles

According to Xinhua News Agency, the concept of “AI-defined vehicles” represents a fundamental shift in how cars are designed and experienced. Zhou Xinyue, an analyst at CCID Consulting’s Automotive Industry Research Center, explained that “the core of AI-defined vehicles is to transform vehicles from passively executing terminals into mobile intelligent agents with autonomous thinking capabilities.” This goes beyond traditional software-defined vehicles, which focused primarily on hardware-software decoupling and over-the-air updates.

Meng Pu, Chairman of Qualcomm China, described two key trends driving this transformation: AI serving people inside the cockpit—whether driver or passengers—and AI providing solutions for the physical world by integrating the vehicle with its surrounding environment during driving. “As AI capabilities continue to be deployed in vehicles,” he said, “the value of cars will evolve from ‘getting people to their destinations’ to becoming intelligent companions that ‘understand, assist, and serve people.’”

From Voice Commands to Multimodal Intelligence

Gou Xiaofei, Senior Vice President of R&D at Li Auto, offered a vivid metaphor for the change: “In the past, voice assistants in the cockpit could basically only ‘listen.’ Now we’re letting them ‘open their eyes’—they can both ‘listen’ and ‘see,’ and also ‘think proactively.’”

The Li Auto L9 Livis, released in May 2026, exemplifies this new generation of AI-powered vehicles. Based on the Snapdragon 8797 platform with local AI computing power, it deploys a multimodal large model on the device side covering vehicle control, navigation, entertainment, lifestyle services, and cross-device integration. A single natural language command can trigger the AI agent to complete complex multi-step tasks like closing windows, playing music, finding a destination, and adjusting cabin ambiance.

Cockpit-Driving Integration: A Structural Revolution

One of the most significant developments highlighted at the summit is the convergence of smart cockpit and driver assistance systems—previously handled by separate departments using different solutions. The new paradigm of “cockpit-driving integration” unifies them through centralized computing, shared computing power, and data interoperability.

BAIC Group Vice President Liu Yu noted that production models including the Arcfox WenDao V9, new Alpha T5, and new Alpha S5 now support memory parking, highway and urban NOA (L2+), and multi-screen displays. “Relying on world models and VLA technology upgrades,” Liu said, “future cockpit-driving integration solutions will continue to iterate, integrating stronger AI capabilities to make intelligent driving safer and more aligned with travel scenarios.”

Democratizing Intelligent Driving

A key theme emerging from the summit is the democratization of advanced driving features. According to CCID Consulting data, L2-level and above assisted driving penetration in China is approaching 70%, while urban NOA has dropped from the premium 300,000+ RMB segment to the mainstream 150,000–200,000 RMB price range. A joint report by the China Consumer Products Quality and Safety Promotion Association and others shows that smart cockpit penetration in new vehicles has reached 73% in China and 58% globally.

Liu Yu emphasized that “equalization of intelligent driving is an inevitable trend in industry development. High-level assisted driving should not be exclusive to high-end models but should become standard across all price segments.” BAIC is specifically targeting the 100,000–200,000 RMB family car segment with cost-effective cockpit-driving integration solutions developed in partnership with Qualcomm, Cheliantianxia, and Zhuoyu Technology.

Policy Support and Ecosystem Collaboration

China’s government has established a comprehensive policy framework supporting this transformation. The New Energy Vehicle Industry Development Plan (2021–2035) specifies “electrification, connectivity, and intelligence” as development directions, while the “Opinions on Deeply Implementing the ‘AI+’ Action” calls for developing intelligent connected vehicles as next-generation smart terminals.

Tao Hailong, Party Secretary and General Manager of SAIC Volkswagen, argued that the industry must move beyond simply competing on features and parameters. “AI-empowered vehicles must shift from a ‘command center’ to an ‘intention center,’” he said, “enabling vehicles to predict user needs and become ‘mobile intelligent lifeforms’ that can sense emotions, predict intentions, and move flexibly across domains.”

To accelerate this vision, Qualcomm, ThunderSoft, Cheliantianxia, Banma Intelligence, Desay SV, Megatronix, and Zhongke Chuangda jointly launched the “Vehicle-side AI Claw Ecosystem Plan,” designed to accelerate large-scale deployment of in-vehicle AI agents through industry-wide collaboration.

What’s Next

As China’s automotive industry accelerates its transition from electrification to intelligence, the convergence of supportive policy, rapid consumer adoption, and intense competitive pressure is creating a uniquely fertile environment for AI-automotive integration. Nakul Duggal, EVP and General Manager at Qualcomm Technologies, noted that Chinese automakers see a faster return on investment cycle for AI investments compared to other global markets, driven by China’s mature digital ecosystem and consumers’ rapid embrace of new technologies.

The key questions going forward center on how these technologies will scale across price segments and how challenges around safety validation, data privacy, and system reliability will be addressed as AI takes on greater decision-making responsibility in vehicles. What is clear is that China’s cars are not just getting smarter—they are learning to understand the people inside them.