Tesla FSD, double doves again? This is not a simple postponement. Recently, Tesla's official website updated the FSD supervision version of the global list of available regions, and China disappeared. You know, as recently as May this year, China also appeared on Tesla's list of available FSD regions. In just three months, from the "selected candidates" in the list, they became "outsiders". The official website explained that the unopened market requires clear approval from local regulations - to be clear, the approval has not yet come down. What is even more exaggerated is that there have been rumors from netizens that Tesla's Shanghai data center has been "empty", and this matter has been forwarded by many big Vs. In response, Tesla China dismissed the rumor to the car house, calling it false news, the data center is operating normally, and the recruitment of assisted driving is also accelerating. Although Tesla clarified that the news was a misinformation, the list was repeatedly added to rumors, which inevitably led to the buyout of FSD and the drumming in the hearts of prospective owners looking forward to FSD entering China. China was also included in the list in May, why was it removed in August? Those owners who have already spent 64,000 yuan to buy FSD, are entitled to support the cause of human Mars? For more exciting videos, look at the timeline from "entering China" to "disappearing" in one or three months on the AutoHome video platform. In May this year, Tesla announced the regions where the FSD supervision version is available, and China was on the list. As soon as the news came out, the market quickly interpreted it as: FSD is finally about to enter China. But there is a key difference here: the FSD supervised version is also L2 assisted driving, and drivers still need to keep an eye on the road and be ready to take over. In April, Tesla's CFO said it was working with Chinese regulators to get approval for a wide rollout in the third quarter. However, Tesla did not say whether the FSD that is expected to be approved is the latest version. So, the “China List” in May is more like an Oolong event, representing Tesla's expectations for opening up the Chinese market, rather than a real commercial “passport”. It can even be said that this update of the list is essentially bringing back the expectations that were too optimistic before and the regulatory reality. By August, Tesla had retooled its page to take China out of the open area. It's also not the first time Tesla has hit a wall in China, where it tried to drive a limited FSD experience in 2025, but the pace of subsequent pushs has been influenced by factors such as regulation and local adaptation. Many Model Y owners who have spent 64,000 yuan on FSD functions have said that they have been "pitted", and some owners plan to sell their cars because they are waiting for FSD to become hopeless. So the question arises: Why is it so hard for FSD to get into China? 2. Why is it so difficult for FSD to enter China? The real core of FSD is not what sensors are installed on the roof, but the data and training capabilities behind it. Tesla's logic is simple: the more the team runs, the more data is collected; the more data is collected, the faster the model is trained; the model is upgraded, and then the vehicle is pushed back through OTA. This set of "Data-Training-OTA" flywheels can operate at high speed in the United States. But when it comes to China, this closed loop has to be picked up again. So over the years, Tesla has been supplementing infrastructure in China: Shanghai data center, local data closed-loop, lingang AI training capabilities, and cooperation with local companies such as Baidu Maps. In May this year, Tesla also recruited a large number of intelligent driving testers in Beijing, Shanghai, Guangzhou, Shenzhen, Chengdu and other cities. These actions show that Tesla is not unprepared, but is trying to turn the FSD into a version that is truly adapted to China's regulatory and road environment. But the problem is, with data, you still have to have computing power. Tesla said that the current local computing power can meet the training needs of the Chinese market, but the specific chip used and the size are not disclosed. Nvidia CEO Huang Jenxun has revealed that they have obtained permission from the US side to export H200 AI chips to China, but have not yet received final approval from the Chinese authorities. This means that Tesla can't replicate the California supercomputing set in China, and he is unlikely to purchase a domestic hashrate card to run his own model. The result is that the backbone model of the United States, which is fed by the global fleet and iterated wildly, is bifurcating from the local model on the Chinese side, which is sandwiched between the two sides and can only be moved by small steps. And the longer it takes, the wider the gap between the two versions. Model Y L III. Strict domestic supervision The review of new energy vehicles by domestic supervision will only become more and more stringent. On August 21, the General Administration of Market Supervision issued a recall announcement of 13 car companies, totaling more than 7.1 million vehicles, a new single-day high. Among them, Tesla's two recalls exceeded 5.71 million vehicles, accounting for nearly 80%, involving issues such as door handles and combined driving assistance. This sends a clear signal: car companies can roll technology, but the string of safety can not be loosened. Assisted driving is no exception. The first step for Tesla to enter China is not to directly move FSD in. The expression "fully autonomous driving capability" used in the Chinese market has now been changed to "Tesla assisted driving". But for supervision, the name change is only the first step, and the real thing to verify is whether this system can work stably and safely in China's complex roads and extreme scenarios. This is precisely where Tesla is more tricky. China's human-vehicle hybrid, takeaway electric vehicle crossing, plugging, unprotected left turn, coupled with complex scenarios such as heavy rain, dense fog, and backlight, are very typical Corner Cases for pure visual solutions. Tesla must produce enough local test data to prove that the pure vision solution can also meet the safety requirements of the Chinese market. If it is further upgraded to L3 in the future, the requirements will only be higher. The mandatory national standards for L3 and L4 autonomous driving systems issued in August 2026 further strengthen the requirements for risk management, safety and security, road testing and so on. This means that Tesla wants to enter China with the L2 supervision version first, and then upgrade to L3 in the future, and the regulatory threshold will only get higher and higher. This is not a separate threshold for Tesla, but a new stage for the entire smart driving industry: technology can be more and more radical, but safety responsibilities and regulatory standards will only become more and more stringent. To summarize: Previously, FSD was a technical myth. But after waiting for a long time, the myth will fade. For owners who have spent money on FSD but have been slow to use it, and for those who have been waiting for FSD to enter China, it is really disappointing that China has disappeared from the list this time. But the deletion of the list does not mean that the FSD will abandon China. After all, China is one of Tesla's most important markets in the world. With Tesla's sales volume and market size in China, it is hard to imagine that it will easily give up the FSD cake. Therefore, the probability of FSD entering China is only a matter of time. As for how long it will take, no one can give an answer now. (By/Automobile House Peng Fei).