Recently, according to multiple media reports including People's Daily Online and Guangming Net, Chongqing investigated the first AI fake advertising case – the queueing to buy car footage was all fake! In the video, the store was packed with people, grabbing cars like grabbing cabbages, with captions "Blind orders are really too hot".
The result was that law enforcement officers found not a single real person and no real transactions; it was purely an AI performance of a hot sales drama.
My God, this is not selling cars? Even hiring actors would be considered troublesome! Intentionally faking facts and concealing the truth to create anxiety of "if you don't buy now it's gone", forcing you to follow the trend and place orders, isn't this naked consumer fraud? Thinking it was done flawlessly, once the regulation checked it, the true face was exposed.
•The trust crack caused by AI-car making is tearing up the entire industry
AI faking is not new in the auto circle, but this recent operation is absolutely outrageous. Have you seen screenshots of "Lei Jun live streaming sales of AITO M9" or "Yu Chengdong standing up for Xiaomi SU7"? The interface layout, character expressions, and bullet screen interactions were highly realistic; if not for official rumor refutation, the public opinion impact would have been huge.
These fake content spreads extremely fast, often covering millions of users within a few hours, bringing immense public opinion pressure to brands.
Even more absurdly, some accounts even let Lei Jun "sell" cars from other brands, and Yu Chengdong "promoted" Xiaomi, making "cross-brand standing" in the AI era completely handled by algorithms.
If the big bosses bringing each other counts as "entertaining memes", then the malicious tampering targeting specific car models is the real tumor shaking the industry foundation. During the release of Voyah Tai Shan X8, an image of "Chairman Lu Fang posing with the new car, trunk turned into 'Man's Wardrobe'" swept the entire internet.
The scene was complete and people looked natural, causing many users to believe it, even asking about the press conference location. The truth was: the original photo was a group photo of other public figures, and after AI fast face swap, scene replacement, and copy modification, it became a "news photo" misleading the market in minutes.
Finally, the Chairman of Voyah had to step out personally to clarify, stopping the spread of rumors.
Similar events have erupted intensively in the past two years; almost all mainstream brands have not escaped. Xiaomi Auto has released official statements 2 to 3 times in the past two years, refuting AI-generated content including fake executive appointments, forged model function copy, etc.;
During the Guangzhou Auto Show, Xpeng Cars encountered AI-generated vulgar exhibition hall videos spread maliciously, with the Legal Department urgently refuting the rumor;
Xpeng GX interior "spy photos" were massively forwarded, seeming like real sneak shots, but the rearview mirror proportion was distorted and interior lines misaligned, typical AI loopholes were spotted at a glance;
Avatr also released a legal statement pointing out that the false publicity images severely infringed on the brand reputation.
From the current industry phenomenon, car circle AI images can be roughly divided into three categories, with hazards increasing step by step:
The first category is the entertaining "wish list images", such as free new car replacement, high points gifts, etc., leaning towards memes, relatively harmless spreading;
The second category is the "half-real images" with realistic details, tampering with configuration, parameters, interior, and prices based on real car models, professional, extremely difficult for ordinary consumers to distinguish;
The third category is the malicious defamation "black PR images", attacking brands through vulgar, distorted, and fabricated content, causing huge damage to reputation.
What truly shakes the industry foundation is those fake-looking real high-fake contents in the middle zone, like Shangjie Z7 winter test official photos. Clearly real car shots, just because of abnormal wheel hub visuals, were directly judged by netizens as "AI generated", and the official had to clarify specifically that it was just a confidentiality design.
Official promotional photos of brands like AITO, Li Auto, Xpeng, etc., are also frequently enlarged by netizens to compare light and shadow, lines, and pick out details. As long as there is a slight visual deviation, they will be labeled "fake photos".
More terrifying than AI faking is the whole industry falling into a weird loop of "proving innocence". In the past, when car companies released official photos, winter test videos, spy photos, users tacitly believed them to be real. Now, official content instead has to accept the whole internet "authentication".
Car companies are forced into a vicious cycle, releasing content – being questioned – urgently refuting – being questioned again. When reality needs repeated self-proof, industry trust has quietly collapsed.
Industry data shows that the judicial appraisal cost of a single AI image has exceeded 50,000 yuan; more than 90% of fake content is spread after more than 10 reposts, the source tracing difficulty is huge; the legal boundary between entertainment memes and commercial defamation is blurred, evidence collection cycle is long, rights protection efficiency is low.
For car companies, whether letting it go or enforcing accountability, they have to pay high costs. This is already not the dilemma of a single brand, but a systemic problem facing the whole industry together.
And this Chongqing case of "enterprise actively using AI to fake queueing" pushes this crisis to a new height; it is the first case of enterprise active faking to mislead consumers as announced by officials.
Today can use AI to fake hot sales, tomorrow can use AI to fake sales volume, reputation, review. Don't always think about going the wrong way to play tricks, relying on false publicity to gain traffic and scam customers, in the end it will only bring stones to break one's own feet.
The essence of this crisis is the collision between digital technology and traditional industrial civilization.
The auto industry is inherently focused on precision, rigor, and safety; users buy real chassis, power, handling, safety performance, not a beautiful picture.
When AI makes text, images, and videos unreliable, real experience and authoritative data are becoming the new trust cornerstone. 16 million new energy vehicles sold in the whole of 2025, penetration rate broke through 50% for the first time; in some months in 2026 even pushed above 55%. The more white-hot market competition, the more likely those companies relying on AI "adding extra scenes" will be backfired.
Fortunately, regulations have already acted. On April 26, 2026, the State Administration for Market Regulation launched a six-month Internet advertising rectification action, explicitly listing "AI generative advertising" as one of the six key points.
Then on May 26, the Ministry of Industry and Information Technology released "2026 Automotive Standardization Work Points", systematically laying out automotive AI standards. According to experts: "AI will become a booster for mobile intelligent agents, but misuse is a rollover accelerator."
•Real products and user reputation are the best advertising
The opposite of "fake queueing" is exactly those players speaking with hard power. When the "hot" forged by AI becomes a joke, the real hotness is happening in the factories and delivery centers of BYD, Deepal, Xpeng, and SAIC.
At the 2026 Beijing Auto Show, BYD booked the entire E3 Hall, directly setting up a "-30°C low temperature flash charging" actual test area and full scenario parking experience area;
Deepal L06 Max limited time starting from 118,900 yuan, all series standard with CATL Golden Bell Battery, even bringing magnetic rheological chassis which only million-level luxury cars have.
When others are still "adding scenes" in the editing room, these brands are already betting with real technology.
Look at intelligent driving again. Among Xpeng Ultra users, the proportion of "opening intelligent driving every day" in the first week of new cars soared to 98.52%, March Ultra model orders increased by 118% month-on-month. He Xiaopeng's original words: "Intelligent driving has changed from showing off skills to a rigid decision point."
SAIC even launched the concept car of "Global First AI Native Car" Jia Yue 07, making the whole vehicle hardware customized for the AI underlying layer.
Even Shangjie, which has been troubled by AI rumors, insisted on releasing original test data after the Z7 winter test controversy, using real parameters to fight back questions.
Industry change has quietly occurred, more and more brands choose to release original data of actual tests, Li Auto and other brands actively strengthen information transparency; China Automotive Technology and Research Center and other authoritative institutions' crash tests, professional review traffic increased by about 30% year-on-year; offline test drives, real human actual tests, live car teardown, etc. verifiable content is more convincing than refined text and images.
Consumers are not stupid either; whose product can withstand real road tests, whose data dares to face third-party teardown, that is who will win the next round of voting.
Big boss speeches are also getting more realistic. In May this year, at the Shenzhen Future Automotive Pioneer Conference, Avatr Chairman Wang Hui spoke directly: "Sales without profit is fake sales."
Wang Lang added a knife: "Rather than exhausting in internal involution, grow in value."
Li Auto's Li Bin gave data: January to April cumulative delivery over 112,000 vehicles, year-on-year growth 71%, operating profit profit for two consecutive quarters.
BYD's Wang Chuanfu's sentence "Building high-end brands is never a one-step process", perhaps is the clearest footnote to the industry.
Pull the lens back to see – 2025 Chinese car exports 6.8 million vehicles, accounting for 18.6% of total sales. In Europe, Chinese share 7.4% surpassed Korea for the first time; in Thailand, Chinese brand share 34.3% firmly No. 1; in Brazil, new energy field already holds half.
Chery exports nearly 1.2 million vehicles to become the "export number one"; SAIC MG delivered 250,000 in Europe, year-on-year growth over 20%. I ask: If Chinese cars cannot present solid technology and products, can they conquer so many markets globally at the same time?
In the era where AI can "create" everything, truth is the longest lasting competitiveness. Just like Dali Lun for Great Wall Motor that heartfelt "Great Wall Cannon" – that is a real emotional expression, no script, no AI rendering, just happened to become the most scarce content for brands now.
When your product can be liked by real users, can be spread out in reality, then your brand itself is the cheapest and most effective advertising. This is the two-way rush towards real human touch.
•Electric EV:
The person in charge of Chongqing Liangjiang New Area Market Regulation Bureau said clearly: "AI is a normal technical tool, but it absolutely cannot become the 'shield' for false publicity."
A series of absurd events are all throwing out the same question: In an era where algorithms can easily copy all visual appearances, what is the most irreplaceable core of the auto industry?
Chinese cars are charging to the world at unprecedented speed, rather than using AI to fabricate "hotness", better to light up the market with products, what do you think?
© 2024 Bitauto.com. All Rights Reserved.
Bitauto (www.yiche.com and www.bitauto.com) is one of the most visited automotive websites in China. We provide detailed info on Chinese car brands available in your area. Please note that the information provided on this website is based on publicly available data and is for reference only. It may not be fully updated or accurate at all times.
