On the morning of 17 September, Yang Guojie, a lawyer at Qianhai Law Firm, was invited to deliver a specialist public legal education lecture entitled “Autonomous Driving Levels and Legal Liability Determination” to lawyers and enterprise representatives.

In his lecture, Yang Guojie began with the classification of autonomous driving and, on the basis of the national standard GB/T 40429, set out the technical characteristics and allocation of liability at each level from L0 to L5. At L0 to L2, the driver is the sole subject of liability. At L3, where an accident occurs after the system is activated, the principle of presumed fault applies, and the operating platform must provide back-end data to establish its innocence, failing which it bears the consequences of failing to discharge the burden of proof. At L4 and L5, where the system takes full control, the traditional human-centred fault liability framework no longer applies, and the subject of liability shifts to the operating enterprise and platform.

▲   Lawyer Yang Guojie, keynote speaker

Referring to the Regulations on the Administration of Intelligent Connected Vehicles of the Shenzhen Special Economic Zone and recent judgments of Shenzhen courts, Yang Guojie focused on the rules by which administrative liability shifts from the driver to the vehicle manager and the operating enterprise. Citing a case from the Shenzhen Intermediate People’s Court, he explained the logic for allocating liability among the parties in scenarios involving defective road facilities, passenger misconduct and third-party fault.

Addressing practical questions of concern to enterprises, Yang Guojie set out the operational path for pursuing remedies: secure evidence immediately after an accident — screenshots, screen recordings, and saved transaction and conversation records, with notarised preservation where necessary; then complain to the platform requesting removal of infringing content and disclosure of the infringer’s information; report in parallel to regulators such as the 12377 reporting centre; and finally resolve the dispute through litigation. He particularly reminded operating enterprises that algorithmic transparency, data retention and security obligations have become compliance baselines, and that whether back-end data can be submitted promptly and completely directly determines how liability is determined in court.

Participants said the lecture was at the forefront of developments and well supported by cases, and offered strong guidance for understanding the legal risks of intelligent connected vehicles and improving corporate compliance systems. Going forward, the parties will continue to follow autonomous driving legislation and judicial practice and jointly promote the sound and regulated development of the intelligent connected vehicle industry.