Our Privacy Statement & Cookie Policy

By continuing to browse our site you agree to our use of cookies, revised Privacy Policy and Terms of Use. You can change your cookie settings through your browser.

I agree

Tesla to roll out limited FSD features in China, boosting domestic smart driving development with DeepSeek integration

CGTN

A Tesla charging station is seen connected to a vehicle in the Changning district of Shanghai, February 24, 2025. /VCG
A Tesla charging station is seen connected to a vehicle in the Changning district of Shanghai, February 24, 2025. /VCG

A Tesla charging station is seen connected to a vehicle in the Changning district of Shanghai, February 24, 2025. /VCG

Tesla on Monday made a long-awaited update to its autopilot software in China to add a city navigation feature that is considered a limited version of its Full Self-Driving (FSD) feature available in the United States.

The U.S. electric vehicle maker, in a notification on its app, said capabilities will include automatic lane-changing depending on speed and route as well as detecting traffic lights at intersections and deciding whether to turn. An in-car camera will also monitor driver attention, it said.

The features are limited due to insufficient data training on Chinese roads and traffic rules, Reuters reported, citing a person with direct knowledge of the matter.

FSD is a suite of driving-assistance technologies developed with generative artificial intelligence to cope with more complicated traffic conditions whereas autopilot handles more routine conditions.

Musk wants to bring full autopilot and FSD systems to China but has said that the effort was proving to be tricky due to technological limitations. The company has postponed an initial goal of end-2024 to this year.

Tesla falling behind

The delays to rolling out a full FSD system in China have left Tesla on the back foot versus Chinese rivals such as Huawei, Xiaomi and BYD, which have launched dozens of EVs with advanced smart-driving software capable of navigating through China's complicated city traffic.

Tesla charges its customers an additional fee of nearly $9,000 to access a limited version of its FSD software in its EVs priced from $32,000.

By comparison, Xiaomi, whose SU7 electric sedan is priced from $29,700, has outsold Tesla's Model 3 in China on a monthly basis and offers an enhanced smart-driving system, including city navigation, for free.

Earlier this month, BYD also started offering advanced autonomous-driving features on most of its models for free, including ones priced as low as $9,555.

Tesla's autopilot update is on par with BYD's mid-tier version of its God's Eye system, which uses advanced computing and lidars to generate 3D images of a car's surroundings to assist navigation around obstacles. It is offered in its premium brand Denza and high-end models of its Dynasty and Ocean series.

Musk has said that Tesla is seeking regulatory approval from Beijing to fully roll out FSD in the country, but China currently only requires automakers to register any level-two update with the industry ministry.

Under industry standards, both autopilot and FSD are categorized as level-two autonomous driving technologies, so under Chinese law, they require human supervision and intervention when necessary.

Domestic development accelerates with DeepSeek integration

Meanwhile, Chinese carmakers are rapidly rolling out moves to enhance in-car experiences and smart driving capabilities, especially after the introduction of the DeepSeek language model. In early February, Chinese car manufacturers, including Geely Auto, IM Motors, Leapmotor and GAC Group, announced deep integrations with DeepSeek's AI model in boosting intelligent driving development.

Leapmotor distinguished itself as the first among new energy vehicle startups to incorporate DeepSeek, following Geely Auto's earlier announcement, which made Geely the first automaker to officially collaborate with DeepSeek.

These integrations aim to enhance features such as multi-turn voice assistant commands, text-to-image generation and active learning. By leveraging advanced central domain controllers equipped with powerful AI chips and optimized computing power, the collaborations seek to refine cockpit interaction logic, bolster natural language understanding and accelerate personalized service responses. This initiative ensures a seamless and efficient experience across various applications, including intelligent cockpits and human-machine interactions.

To date, over a dozen automakers and brands – including BYD, Geely Auto, Voyah, Baojun, Dongfeng Motor, IM Motors, Leapmotor and GAC Group – have announced comprehensive integrations with DeepSeek's AI model.

(With input from Reuters)

Search Trends