The reason "who is making rushed cars" is so hard to answer is that "rushed development" has been treated as a label to fling at others rather than an engineering judgment that can actually be tested.

The yardstick for a rushed car is not speed, but whether the validation process has been stripped down.

The definition offered by People's Daily points to "excessively compressing the R&D cycle and cutting the real-vehicle validation process." Lu Fang, chairman of Voyah, put it more bluntly: the only yardstick for judging whether a car is rushed is not the length of time, but whether the validation process is complete. So the real dividing line is not "fast" itself.

The problem arises when "fast" turns from one company's engineering capability into a goal the entire industry feels compelled to chase. R&D pace is then no longer determined by engineering principles, but pushed backward by launch windows, marketing milestones, and competitors' release dates.

First set the launch date, then let testing work backward from the schedule. The launch event has already been prepared, channels have been laid out, and orders have been taken. At this point, if testing uncovers problems, anomalies that have not yet been clarified may simply be left for handling after launch.

The fact that more than thirty automakers were summoned for discussions precisely shows that regulators do not judge this to be an isolated phenomenon, but a systemic pace deviation. GAC can confidently say that it "resolutely does not build rushed cars," citing 12 major validation items and more than 1,500 validation sub-items—but this only proves that GAC is not on the list; it does not prove that the list is empty.

In the first half of the year, 542 new cars launched in clusters, but only 75 were truly new platforms.

The numbers themselves speak. In the first half of 2026, a total of 542 new cars were launched domestically, averaging more than three per day, but only 75 models were truly based on all-new platforms, with the vast majority of the rest being annual facelifts and derivative versions.

In September alone, more than 50 new cars were launched in a single month, and on September 23 alone, nine cars debuted at the same time.

Zeng Yuqun, chairman of CATL, also gave a broader figure at the Power Battery Conference: since the beginning of this year, more than 600 new cars have been launched domestically, averaging close to three per day. The development cycle has been compressed from about 60 months in the fuel-vehicle era to 24 months, and many new energy models have even been shortened to within 18 months.

Being fast is not a sin in itself, but when an industry uses 18 months to cover a calibration window that previously required three to five full winters to complete, what is omitted does not disappear; it is merely transferred to users.

The reason "two winters and one summer" has become the toughest rule in the new regulations is precisely that winter comes only once a year, and the physical window for cold-climate calibration cannot be compressed by capital or marketing rhythm.

Changing a bumper and daring to call it all-new: reshelling and oil-to-electric conversion are the hardest-hit areas.

When users ask "how many rushed cars are there with just a changed bumper and badge," the answer lies in the way the word "all-new" has been abused.

A true all-new generation involves simultaneous updates to the platform, chassis, and powertrain architecture, with substantive changes in body dimensions and core hardware; whereas a large number of so-called "new models" merely redesign the front grille, light clusters, and bumper, with the wheelbase unchanged, the chassis carried over, and no underlying breakthrough in the electric drive or powertrain.

No matter how big the exterior changes to a reshelled car, the foundation is still the old one. More alarming than reshelling are hastily launched "oil-to-electric" conversions: the battery is not laid flat on a dedicated pure electric chassis, but is squeezed into the original positions of the driveshaft and fuel tank wherever possible, resulting in a raised rear floor, insufficient battery ground clearance, and an increased risk of battery damage after a collision.

The problem with such models is not "electric" itself, but that the chassis was not designed for batteries and high-voltage systems from the start, and the margins for structural safety and thermal management were already limited at the project approval stage.

Simulation can calculate fatigue curves, but not the real pace of rust eating through a chassis.

The gap between laboratory simulation and real road conditions is not a question of "whether simulation is useful," but of "what simulation can replace and what it cannot."

Simulation is indeed efficient in validating design directions in the early stages of a project and estimating signal quality at the bottom of chips. Equivalent load fatigue of 300,000 kilometers can be run in a simulation environment in days or even hours. But material performance in the real world is not completely consistent with theoretical parameters. Frequently moving mechanical mechanisms such as door hinges and seat rails also face nonlinear random factors such as wear, lubrication degradation, and dust intrusion.

More critically, for items that directly concern life, such as crash safety, structural fatigue, and battery thermal runaway, the physical testing hurdle cannot be omitted.

Some industry practitioners revealed that for a door handle life test of 10,000 cycles, simulation ran 8,000 cycles, while the real vehicle was only tested 2,000 times. Simulation can calculate fatigue curves, but it cannot calculate the batch failure pace of air spring seals in Chongqing's 40-degree heat, nor can it calculate the peeling of anti-corrosion coatings from the inside after repeated freeze-thaw cycles with road deicing salt.

The validation costs saved do not vanish into thin air; they are merely deferred into repair orders, insurance claims, and used-car residual values.

Overseas reputation is a mirror: strict regulation of rushed cars is an act of responsibility toward global users.

The scale of China's automobile exports no longer allows the logic of "sell first and talk later" to continue.

Tests by Russia's Lamark Auto Anti-Corrosion Center covering 26 regions showed that among 17 tested Chinese cars, five developed chassis rust-through in just two years, with anti-corrosion coating thickness only half that of German cars.

A survey by the Russian newspaper Kommersant gave an average failure cycle of about three years for Chinese cars, while German cars can run stably for more than six years.

These figures are certainly amplified by usage environments and climate conditions, but it is precisely this amplification effect that has laid bare, all at once, the validation shortcomings that are not easily exposed under mild road conditions in China. Strict regulation of rushed cars benefits Chinese users most directly, but equally important is that it determines how Chinese automobiles will ultimately be remembered in the global market: as "cheap goods" that rely on price for volume and are rusty after a few years, or as a reliable choice that can withstand the test of time even in extreme environments.

Regulation raising the market access threshold from "meeting national standards is enough" to "sufficient validation is required for access" is essentially making "complete validation" once again something that cannot be bypassed with marketing rhetoric.

Opposing haste is not opposing efficiency, but opposing treating validation as a cost item that can be negotiated. For Chinese automobiles to go global, the first thing is to ensure that every car that drives out can withstand others looking back at it five or ten years later.