Beijing pilots dedicated insurance for intelligent driving as smart driving enters the ‘accounting’ stage

What Beijing did

According to TMTPost, it has been reported that Beijing on March 29 launched a pilot commercial insurance product specifically for intelligent, connected new-energy vehicles. The move is more than a product tweak; it signals a shift from a pure technology race to building the liability and financial infrastructure needed for large‑scale deployment. The pilot is designed to extend existing new‑energy vehicle policies to cover scenarios and losses tied to L2–L4 automated driving systems — from driver‑assist incidents to conditional or no‑driver modes.

Liability and software‑defined cars

The core problem isn’t sensors or compute alone — it’s who is legally responsible when software, not a human, is effectively driving. Under traditional car insurance the default assumption is “human driver”; but L3 and above mix human and machine control and make fault attribution complex. It has been reported that an insurance executive involved in product design said insurers now must “redefine who is driving,” meaning risk logic must be rewritten to include algorithms, OTA updates and sensor performance. Reportedly, the new product will cover some “soft‑and‑hardware” loss types and aim for faster payouts to protect victims and shorten painful liability fights.

From uncertainty to pricing

Insurance turns unknowns into priced risk. As insurers start to ingest operational data — algorithm versions, system stability and failure modes — they can differentiate systems’ risk profiles and reflect that in premiums. That will make smart‑driving capability a financial variable as well as a product feature: more stable systems should attract lower premiums, while volatile systems will cost more to insure. Amid global tensions over chips and sensors — including U.S. export controls and supply‑chain uncertainty — those technical and geopolitical risks may also be folded into pricing. The question now is not just whether automakers can build autonomous features, but whether there is a whole ecosystem that can shoulder responsibility and put a price on it.