AI-Generated List Sparks EV Industry Crisis in China

An AI-generated list of automakers has triggered a full-blown misinformation crisis across China's electric vehicle industry, forcing 8 major brands — including Tesla, BYD, and NIO — to issue public denials after false rumors spread that they were under government investigation for secretly reducing battery capacity through over-the-air software updates.

AITO (问界), the Huawei-backed EV brand, became the latest company to issue a formal statement on May 9, calling the allegations 'completely fabricated' and confirming that all its OTA (over-the-air) update content has been legally filed and approved by regulators.

Key Facts at a Glance

• 8 major EV brands have now issued public denials of the rumor, including Tesla, BYD, XPeng, Li Auto, NIO, GAC Aion, Zeekr, and AITO
• The crisis originated from an unverified report claiming 8 automakers were 'summoned' by regulators over battery-locking practices
• Users turned to AI chatbots to identify the companies — and the AI fabricated a list covering nearly every major EV brand in China
• No regulatory body has publicly confirmed which companies, if any, were actually investigated
• AITO has stated it is gathering evidence and will pursue legal action against those spreading the rumors
• The incident highlights growing risks of AI-generated misinformation in financial and industrial contexts

How an AI Hallucination Spiraled Into an Industry Crisis

The chain of events began on May 8, when an unverified claim surfaced online alleging that 8 new energy vehicle manufacturers had been 'summoned' by Chinese regulators over so-called 'battery locking' — the practice of using OTA software updates to quietly reduce a vehicle's usable battery capacity. The claim also alleged that 3 of those companies were under formal investigation for regulatory violations.

Critically, no government agency published an official list of the companies involved. This information vacuum created the perfect conditions for speculation to fill the gap.

With no authoritative source to reference, curious social media users did what has become increasingly common in 2025: they asked AI chatbots to identify the companies. The AI tools, lacking access to verified information about the case, did what large language models often do when faced with insufficient data — they hallucinated. The chatbots generated plausible-sounding lists that included virtually every major new energy vehicle brand operating in China.

Social Media Amplification Turned Speculation Into 'Fact'

Once the AI-generated lists began circulating on platforms like Weibo, Douyin, and various WeChat groups, the rumor took on a life of its own. Self-media accounts and influencers amplified the unverified information, presenting the AI-fabricated lists as though they were leaked government documents.

The speed of amplification was remarkable. Within 24 hours, the rumor had become so widespread that companies felt compelled to respond individually. The pattern of corporate responses unfolded rapidly:

• XPeng issued an early denial, distancing itself from the allegations
• Tesla China followed with a statement rejecting any connection to the claimed investigation
• Li Auto and NIO released clarifications on the same day
• BYD, China's largest EV maker by volume, publicly denied involvement
• GAC Aion and Zeekr added their denials to the growing list
• AITO became the 8th brand to formally respond on May 9

The irony is hard to miss: a rumor about 8 companies ultimately prompted exactly 8 companies to deny it, inadvertently lending a false sense of credibility to the original claim's specificity.

The 'Battery Locking' Controversy Explained

To understand why this rumor gained traction so quickly, it helps to understand the underlying consumer anxiety it exploited. Battery locking, known in Chinese EV circles as '锁电,' refers to the alleged practice of automakers using OTA updates to restrict the usable portion of an EV's battery pack.

Manufacturers sometimes adjust battery management parameters to extend battery lifespan or address safety concerns. However, consumers have long suspected that some companies use these updates to quietly reduce range — effectively degrading a product they have already purchased.

This is not a uniquely Chinese concern. Tesla faced similar controversies globally, including a $1.5 million settlement in Norway in 2021 over allegations that a software update reduced battery capacity and charging speeds on older Model S vehicles. The practice touches on fundamental questions about:

• Consumer rights — Do buyers own the full capacity of the battery they paid for?
• Software transparency — Should OTA updates that affect performance require explicit consumer consent?
• Regulatory oversight — How should governments monitor software changes that affect product specifications?
• Right to repair — Does manufacturer control over software constitute an unfair restriction on ownership?

These unresolved tensions made the rumor particularly believable and emotionally resonant for Chinese EV owners.

AI Misinformation Poses Growing Risks to Markets and Industries

This incident serves as a stark case study in how AI-generated misinformation can disrupt real-world industries. Unlike traditional misinformation, which typically requires a human actor to deliberately fabricate details, this crisis was partly generated by AI systems that were simply trying to be helpful.

The pattern is increasingly familiar to researchers studying AI risks. When users ask a large language model a question it cannot answer with verified information, the model often generates a plausible but entirely fictional response. In low-stakes contexts, these hallucinations are a nuisance. In high-stakes contexts — involving publicly traded companies, regulatory investigations, or consumer safety — they can cause real economic and reputational damage.

Compared to the well-documented cases of AI-generated deepfakes or synthetic media, this type of misinformation is arguably more insidious because it appears as factual text rather than manipulated imagery. Users often trust text-based AI outputs more readily than they trust suspicious images or videos.

The Chinese EV industry is particularly vulnerable to this dynamic for several reasons:

• High public interest — China is the world's largest EV market, with over 10 million new energy vehicles sold in 2024
• Intense brand competition — Rumors can shift consumer preferences rapidly in a crowded market
• Regulatory opacity — Government investigations are often conducted without public disclosure, creating information vacuums
• AI adoption rates — Chinese consumers are among the world's most active users of AI chatbots, including Baidu's Ernie Bot, Alibaba's Qwen, and others

What This Means for the Global AI Ecosystem

For Western observers, this incident carries implications that extend well beyond China's automotive sector. It demonstrates a failure mode that any market or industry could experience as AI chatbot usage grows globally.

Consider the parallels: in the United States and Europe, consumers increasingly use ChatGPT, Google Gemini, Claude, and other AI assistants to research companies, interpret news, and make purchasing decisions. If a similar information vacuum emerged around a regulatory action involving, say, major U.S. tech companies or European automakers, the same dynamic could easily repeat itself.

The incident also raises questions about AI platform responsibility. Should chatbot providers implement safeguards against generating speculative lists of companies in regulatory or legal contexts? OpenAI, Google, and Anthropic have all invested heavily in reducing hallucinations, but the problem remains far from solved — particularly for queries about recent events where training data is limited.

For businesses, the takeaway is clear: corporate communications teams need AI-specific crisis response protocols. The speed at which AI-generated misinformation can spread means that traditional 24-48 hour response windows may be far too slow.

Looking Ahead: Regulation and Responsibility

This episode will likely accelerate several trends already underway in both China and the West.

First, expect Chinese regulators to increase scrutiny of AI-generated content on social media platforms. China's existing regulations on AI-generated content, including mandatory labeling requirements that took effect in 2024, may be expanded to cover scenarios where AI outputs are shared as factual claims about specific companies.

Second, the automotive industry globally may push for clearer OTA update disclosure standards. Whether or not any company was actually investigated in this case, the underlying consumer concern about battery locking is legitimate and unlikely to disappear.

Third, this incident will fuel ongoing debates about AI liability. If an AI chatbot generates a false claim that damages a company's reputation or stock price, who bears responsibility — the AI provider, the user who shared the output, or the platform where it was amplified?

As AI tools become embedded in everyday information consumption, the line between a helpful assistant and an unwitting rumor mill grows thinner. The Chinese EV industry just learned that lesson the hard way — and it will not be the last industry to do so.