China Writes the Rules for Global Autonomous Driving: How the ADS GTR Signals a New Era of Technical Leadership

On a Tuesday afternoon in late June, inside a conference hall on the shores of Lake Geneva, something quietly historic happened. Representatives from China, the European Union, the United Kingdom, the United States, Canada, and Japan gathered under the auspices of the United Nations Economic Commission for Europe's World Forum for Harmonization of Vehicle Regulations—better known as WP.29—and voted to adopt the Automated Driving System Global Technical Regulation, or ADS GTR.

The vote was not close. It was not controversial. It was, in the dry language of international bureaucracy, a consensus. But consensus at this scale is never accidental. It means six of the world's most economically consequential regions—regions that have spent the last decade racing each other toward autonomous driving supremacy, often with incompatible rules, conflicting standards, and mutual suspicion—have agreed to play by the same rulebook.

That rulebook now exists. And China didn't just sign it. China helped write it.

For years, China's role in global technical standards was largely reactive: study what the West produces, translate it, adapt it. The ADS GTR represents something different. The Chinese delegation—led by the Ministry of Industry and Information Technology, with participation from Baidu, Huawei, SAIC, and a network of national research institutes—submitted dozens of technical proposals during the drafting process, shared real-world testing data from closed-course, public-road, and vehicle-infrastructure collaborative trials, and advocated for specific safety baselines and validation frameworks that now sit at the core of the global regulation.

When the final vote was called on June 26, 2026, China's fingerprints were not just on the document. They were in it.

This is the story of how that happened. It is not a story of sudden triumph. It is a story of a decade-long accumulation—of policy bets, commercial expansion, technical iteration, and international maneuvering—that transformed China from a follower in autonomous vehicle regulation into a co-author of the world's first global standard. And it is a story about what happens when a technology that was supposed to be a consumer convenience becomes, instead, a matter of national technical diplomacy.

Phase 1: The Long Shadow of the Starting Line (2015–2020)

The modern Chinese autonomous driving story begins not in Beijing or Shenzhen, but in the strategic documents that emerged during the 13th Five-Year Plan period (2016–2020). In 2015, the State Council published the "Made in China 2025" roadmap, which elevated intelligent connected vehicles to one of ten priority sectors. The goal was explicit: by 2025, China would possess the core technologies and industrial capacity for autonomous driving at Levels 3 and 4.

At the time, this ambition seemed almost absurdly ambitious. In 2015, China's automotive industry was still dominated by foreign joint ventures. Domestic brands were struggling to compete on quality, let alone on software-defined intelligence. The concept of a "smart city" was largely theoretical. And the regulatory environment for autonomous vehicles was a patchwork of local experiments with no national framework.

But the policy signal created the market. Companies like Baidu (which had been running its Apollo open autonomous driving platform since 2017), Pony.ai (founded in 2016 by two former Baidu engineers, James Peng and Lou Tiancheng), and Momenta (founded in 2016 by a group of Microsoft Research Asia alumni) emerged during this period, funded by a wave of venture capital that saw autonomous driving as the next frontier after mobile internet.

The early years were not glamorous. They were defined by closed-course testing—vast proving grounds in places like Chongqing and Beijing where autonomous vehicles navigated pre-programmed scenarios under controlled conditions. The Chinese government, cautious about safety and liability, required human safety drivers in all public road tests. Permits were issued city by city. The first official autonomous driving test license in Beijing was granted in 2018. It covered just a handful of vehicles and a carefully mapped route.

But the policy framework was already learning. In 2017, the Ministry of Industry and Information Technology (MIIT) and the Ministry of Public Security jointly issued the "Guidelines for the Road Testing of Intelligent Connected Vehicles," establishing a national framework for local test permits. The document was not particularly permissive—it required extensive safety documentation, insurance coverage, and real-time monitoring—but it created a common language. Cities began competing with each other to attract autonomous driving companies. Beijing, Shanghai, Guangzhou, and Shenzhen rolled out their own test zones, each with slightly different rules, creating a natural laboratory of regulatory experimentation.

By the end of 2019, China had issued over 250 autonomous driving test licenses across 20 cities. The total distance covered in public road tests exceeded 3 million kilometers. It was small by today's standards, but it was a foundation. And it was happening while the United States, through companies like Waymo and Cruise, was moving faster and more aggressively in terms of actual deployment.

The consensus in 2019 was that China was behind. That consensus was not wrong. But it was also not the whole story. China was building something the US was not: a regulatory scaffolding that could support large-scale deployment when the technology matured. The US approach was largely market-driven, with companies pushing regulators to keep pace. The Chinese approach was state-directed, with regulators designing the track before the cars could run on it.

That difference would matter enormously in the years to come.

Phase 2: The Inflection Point (2020–2023)

The COVID-19 pandemic was, paradoxically, a catalyst for autonomous driving in China. In 2020, as human drivers were restricted by lockdowns, Baidu's Apollo Go deployed its first fully driverless robotaxi fleet in Changsha, Hunan Province. The move was partly a response to the pandemic's labor shortages, but it also demonstrated something that skeptics had doubted: autonomous vehicles could operate safely in real urban environments without a human safety driver in the front seat.

The pandemic also shifted government priorities. The "New Infrastructure" initiative, launched in 2020, explicitly included intelligent connected vehicles and smart roads as core components. Billions of yuan flowed into 5G infrastructure, edge computing nodes, and roadside sensing equipment that would eventually enable vehicle-to-infrastructure (V2I) communication. The theory was straightforward: autonomous driving would not work as a vehicle-only problem. It required an ecosystem. China, with its state capacity to coordinate across telecommunications, automotive, and urban planning sectors, was uniquely positioned to build that ecosystem.

By 2021, the regulatory landscape had shifted dramatically. Beijing authorized fully driverless commercial robotaxi services in designated zones. Guangzhou and Shenzhen followed. Baidu's Apollo Go began operating in Wuhan—a city that would become the world's largest robotaxi market—with an aggressive expansion strategy that saw hundreds of vehicles deployed across the city. The rides were cheap, often 30–50% below the cost of a human-driven Didi, and they were real. Passengers ordered them through apps, waited for a car to arrive with no driver inside, tapped a button to unlock the door, and rode to their destination.

The Chinese robotaxi industry was born during this period. But it was also shaped by a different logic than its American counterpart. In the US, Waymo pursued a premium, safety-first approach, expanding slowly and methodically. In China, the playbook was closer to that of the ride-hailing wars of the mid-2010s: deploy fast, subsidize heavily, capture market share, and worry about unit economics later. Baidu's Apollo Go completed its 1 millionth ride in 2022. By early 2023, it was operating over 1,000 fully driverless vehicles across 15 cities.

The regulatory framework kept pace, but with a distinctive Chinese character. The government established a national-level safety assessment system for autonomous driving systems, requiring companies to pass extensive testing on closed courses before receiving public road permits. The assessments were rigorous but also standardized, creating a common bar that all companies had to clear. This was different from the US approach, where regulation was largely a matter of state-by-state patchwork, with no national safety standard.

During this period, Chinese companies also began internationalizing. WeRide secured autonomous driving permits in the UAE, France, and Switzerland. Pony.ai expanded into Singapore and the Middle East. Baidu announced plans for Dubai and Europe. The strategy was clear: the Chinese domestic market was large enough to build the technology, but global deployment would be necessary to amortize the enormous R&D costs.

But the period was not without setbacks. The US-China tech rivalry intensified. American export controls on advanced semiconductors began to bite, affecting the compute hardware necessary for autonomous driving. And the commercial reality of robotaxi economics remained brutal: companies were burning billions of yuan with no clear path to profitability.

By late 2023, the industry had reached a crossroads. The technology worked. The regulatory framework existed. But the business model was still uncertain. And the question of global standards—how to sell Chinese autonomous vehicles in foreign markets with different rules—was becoming urgent.

Phase 3: Acceleration and the International Chessboard (2023–2025)

The years 2023 to 2025 marked the period when China's autonomous driving industry scaled commercially and engaged internationally with unprecedented intensity.

On the domestic front, the numbers grew rapidly. By 2025, the penetration rate of new vehicles equipped with advanced driver-assistance systems (ADAS) in China had exceeded 60%. This was not fully autonomous driving, but it was a massive on-ramp: millions of Chinese consumers were becoming comfortable with vehicles that could steer, brake, and accelerate autonomously in specific conditions. The psychological barrier to full autonomy was being lowered by increments.

The robotaxi fleet also expanded. Baidu's Apollo Go surpassed 250,000 weekly trips by mid-2025, outpacing Waymo's US operations. Pony.ai reached over 3,000 robotaxis by the end of 2025, backed by a strategic partnership with Toyota. WeRide's fleet exceeded 1,000 vehicles and diversified into robobuses, robovans, and robosweepers. The economics were still challenging, but the operational experience was accumulating at scale.

The capital markets also opened. WeRide and Pony.ai both completed Hong Kong IPOs in 2025, joining the HKEX's growing roster of technology and robotics listings. The offerings were not spectacularly profitable—the companies were still losing billions—but they provided crucial capital for fleet expansion and international growth. Hong Kong's capital markets raised approximately HK$166.8 billion ($21.28 billion) in the first five months of 2026 alone, a dramatic surge from the previous year, suggesting that global investors were willing to bet on Chinese autonomous driving companies despite geopolitical tensions.

But the most significant development during this period was the international regulatory engagement. The UN WP.29 process on autonomous driving had been ongoing since 2019, but it accelerated dramatically after 2023. In November 2023, WP.29 adopted a formal proposal for the regulatory approach to Automated Driving Systems, including the collaborative development of a UN GTR and a UN Regulation by 2026. The proposal established an Informal Working Group (IWG) on ADS, with a bureau composed of representatives from Canada, China, the European Commission, Japan, the United Kingdom, and the United States.

China's delegation was not a passive participant. The Chinese representative to GRVA (the Working Party on Automated/Autonomous and Connected Vehicles) served as Co-Chair of the IWG on ADS, alongside Japan. This was a formal recognition of China's technical contributions and diplomatic influence in the process. From this position, China helped shape the "safety case" approach that the GTR would eventually adopt—an approach that requires manufacturers to document and demonstrate that their systems are free from unreasonable risk, rather than simply passing a fixed set of predefined tests.

The safety case approach was critical. It aligned with the Chinese domestic regulatory philosophy, which was moving toward a lifecycle-based safety management system rather than a static type-approval model. The Chinese government, through MIIT, organized domestic research institutes and industry partners to contribute technical proposals and share testing data. The data came from China's unique ecosystem: millions of kilometers of testing in dense urban environments, in complex traffic conditions with pedestrians, scooters, and chaotic intersection behavior that is characteristic of Chinese cities.

This data was valuable to the global process. The GTR needed to account for real-world driving complexity, and China's testing environments provided some of the most challenging scenarios in the world. When the draft GTR was finalized in early 2026, it incorporated performance requirements and validation frameworks that reflected lessons learned from Chinese testing programs.

By early 2026, the pieces were in place. The domestic regulatory framework was mature. The international engagement was deep. The commercial operations were scaling. And the UN vote was scheduled for June.

Phase 4: The Vote and the New Equilibrium (June 2026)

June 2026 was a remarkable month for China's autonomous driving industry.

On June 16, Baidu's Apollo Go received a Level 4 autonomous driving permit in Switzerland, becoming the first Chinese company authorized to operate fully driverless commercial robotaxi services in Europe. The permit was not just a commercial milestone; it was a regulatory validation. Switzerland, with its reputation for precision and safety, does not issue such permits casually. The fact that Baidu cleared the bar signaled that Chinese autonomous driving systems could meet the most demanding international standards.

On June 25, the ADS GTR was formally adopted at the WP.29 session in Geneva. The regulation sets core technical requirements for automated driving system products across their entire lifecycle: safety management processes, product safety documentation, end-to-end testing and validation frameworks, post-deployment vehicle safety monitoring, and audit and evaluation methods. It applies to vehicles of categories M1–M3 (passenger vehicles), N1–N3 (goods vehicles), and L6–L7 (light quadricycles). It covers Levels 3, 4, and 5 of automation, with no operational design domain (ODD) restrictions—meaning the requirements apply to autonomous driving systems regardless of whether they operate on highways, urban streets, or closed campuses.

The regulation was co-sponsored by China, Japan, the EU, the UK, the US, and Canada. It was the result of a decade of work that began with the WP.29 Informal Working Group on Intelligent Transport Systems/Automated Driving in the mid-2010s, evolved through the FRAV (Functional Requirements for Automated and Autonomous Vehicles) and VMAD (Validation Methods for Automated Driving) working groups, and culminated in the integrated GTR/UN Regulation package.

For China, the adoption was a triple victory.

First, it validated the Chinese regulatory approach. China's own mandatory national standard for automated driving systems, which had been developed in parallel with the GTR and was finalized in early 2026, fully aligns with the core technical content of the global regulation. But the Chinese standard goes further, refining requirements for Level 3 and Level 4 systems in light of China's specific industrial conditions and regulatory needs. It introduces standardized test scenarios built on the "multi-pillar" approach, which combines simulation, closed-course testing, and public-road validation. The alignment between the domestic standard and the GTR means that Chinese manufacturers will face no additional regulatory burden to sell their vehicles in GTR-compliant markets. This is a huge competitive advantage.

Second, the GTR adoption cemented China's diplomatic position in the global automotive standards ecosystem. The automobile industry has long been dominated by European, Japanese, and American technical standards. UN Regulations, historically, were largely written by Western and Japanese experts. The fact that China now co-sponsors, co-chairs, and co-authors the most significant new automotive regulation in decades is a symbolic shift that reflects a deeper structural change: the center of gravity in automotive innovation is moving eastward.

Third, the timing was commercially auspicious. The GTR adoption came just as Chinese autonomous driving companies were preparing for massive international expansion. WeRide and Pony.ai had already established beachheads in Singapore, the Middle East, and Europe. Baidu had the Swiss permit and was eyeing Germany and the UK. Momenta, preparing for its $1 billion Hong Kong IPO at a roughly $9 billion valuation, had partnerships with Mercedes-Benz, Uber, and Grab for international robotaxi deployments. The GTR provided a common regulatory baseline that would reduce the compliance costs and accelerate the timeline for these expansions.

The vote was not the end of a process. It was the beginning of a new one. The GTR is a technical regulation, not a law. Each signatory country must transpose it into domestic legislation. China has already done this, with its national standard undergoing final approval procedures. The EU, Japan, and the US will follow over the next 2–3 years. But the framework is now set. The world has a common language for autonomous driving safety, and China helped define it.

Behind the Numbers: A Market in Motion

The ADS GTR is not just a regulatory document. It is the capstone on a market that has already achieved remarkable scale.

Consider the Chinese robotaxi fleet as of mid-2026:

- Baidu Apollo Go: Over 1,000 fully driverless vehicles operating in 15+ cities, with weekly trip volumes exceeding 250,000. The service in Wuhan operates 24/7 in most zones and has become the world's largest commercial robotaxi deployment.

- Pony.ai: Approximately 3,000 vehicles across Beijing, Guangzhou, Shanghai, and Shenzhen, with partnerships for expansion in Dubai, Qatar, and Singapore.

- WeRide: Over 1,000 vehicles, including robotaxis, robobuses, robovans, and robosweepers. The company operates in Guangzhou, Beijing, and internationally in the UAE and Europe. Its robobuses in Guangzhou Nansha accept standard transit cards and require no app registration—making them the most accessible autonomous transport option for tourists.

- Momenta: Primarily a technology provider rather than a fleet operator, Momenta supplies autonomous driving "brains" to 24 global automakers, including SAIC, GAC, Chery, FAW, Geely, and Great Wall. Its revenue reached RMB 2.41 billion in 2025, with gross margins jumping to 71.6%. The company is preparing a Hong Kong IPO that could raise up to $1 billion.

The market size is equally striking. According to industry forecasts, China's autonomous vehicles market is projected to grow from $22.84 billion in 2025 to $218.95 billion by 2034, a compound annual growth rate of 28.5%. This is not a niche technology sector. It is becoming a central pillar of the automotive industry.

The penetration of advanced driver-assistance systems in new vehicles sold in China has already crossed 60%. This is a critical mass effect: as more consumers experience partial autonomy, the demand for full autonomy grows. The regulatory framework, now aligned globally, removes the final barrier to large-scale commercialization.

The Competitive Landscape: China vs. the West

The global autonomous driving race is often framed as a China vs. US competition, with Waymo and Tesla on the American side and Baidu, Pony.ai, and WeRide on the Chinese side. This framing is not wrong, but it oversimplifies the reality.

Waymo, a subsidiary of Alphabet, remains the global leader in terms of safety metrics and technical sophistication. Its operations in Phoenix, San Francisco, and Los Angeles have accumulated millions of miles of fully driverless experience. Waymo's vehicles are conservatively programmed, cautious in complex scenarios, and backed by one of the world's most advanced simulation and validation infrastructures. By the numbers, Waymo still has the best safety record per mile driven.

But Waymo has struggled with scalability and commercialization. Its expansion has been slow, limited by regulatory barriers in the US and by the enormous capital requirements of scaling a fleet. The company has completed fewer total trips than Baidu's Apollo Go, despite starting earlier and operating in more permissive regulatory environments.

Tesla represents a different approach entirely. Tesla's Full Self-Driving (FSD) system relies on a vision-only architecture (no LiDAR, no high-definition maps) and uses a massive fleet of consumer vehicles as data collection nodes. In early 2026, Tesla received partial approval for limited FSD trials in China, with full approval anticipated in late 2026 or early 2027. Tesla's approach is riskier—vision-only systems have well-known limitations in adverse weather and edge cases—but if successful, it could scale far more rapidly than robotaxi fleets because it leverages existing consumer vehicles rather than purpose-built taxis.

The Chinese companies, meanwhile, have pursued a hybrid strategy that combines elements of both approaches. They operate purpose-built robotaxi fleets for commercial deployment (like Waymo) but also supply autonomous driving technology to mass-market consumer vehicles (like Momenta's partnerships with automakers). This dual-track approach provides revenue diversification and data scale.

The regulatory landscape is where the divergence becomes most stark. The US has no national autonomous driving standard. Regulation is a patchwork of state laws, NHTSA guidelines, and local permits. California's approach is different from Arizona's, which is different from Texas's. This fragmentation creates compliance costs and operational complexity that Chinese companies, operating under a single national framework, do not face.

The GTR begins to address this fragmentation. If the US transposes the GTR into federal standards—a process that is expected but not guaranteed—American companies will face a more standardized environment. But the transposition will take years. In the meantime, Chinese companies benefit from a domestic framework that is already aligned with the global standard.

What Comes Next: Implementation and the Road Ahead

The adoption of the ADS GTR is a milestone, not a destination. The real work—implementation—is just beginning.

For China, the immediate priority is to finalize and promulgate its domestic mandatory standard, which is already in the final approval stages. The standard will incorporate the GTR's core requirements while adding provisions tailored to China's specific conditions: clearer safety baselines for different automation levels, stronger user training and disclosure mechanisms to prevent misuse and over-reliance, and the standardized test scenarios based on the multi-pillar approach.

For Chinese companies, the GTR opens international markets that were previously fragmented by incompatible regulations. A vehicle certified under the Chinese national standard, which aligns with the GTR, will have a smoother path to certification in the EU, Japan, and eventually the US. This reduces the cost and time of international expansion.

But challenges remain. The GTR covers safety management and validation, but it does not address data sovereignty, cybersecurity, liability, or insurance—all of which are governed by national laws that remain highly divergent. A Chinese autonomous vehicle operating in Germany will still need to comply with German data protection laws, German liability frameworks, and German insurance requirements. The GTR does not solve these problems. It solves the technical safety baseline problem. The rest is still a matter of bilateral negotiation.

Moreover, the technology itself continues to evolve. The GTR is designed to be technology-neutral—it applies to LiDAR-based systems, vision-based systems, and hybrid systems alike. But as artificial intelligence models become more powerful and as "end-to-end" neural network approaches (which replace traditional rule-based planning with learned behaviors) become more prevalent, the validation frameworks may need to adapt. The GTR's safety case approach is well-suited to this evolution because it focuses on outcomes rather than methods, but regulators will need to develop new expertise to evaluate AI-driven systems that are not easily interpretable.

Finally, there is the geopolitical dimension. The US-China tech rivalry has already affected the semiconductor supply chain, with American export controls limiting Chinese access to advanced AI chips. Autonomous driving is highly dependent on AI compute, and any further restrictions could slow the development of Chinese systems. The GTR provides a common technical framework, but it does not resolve the underlying geopolitical tensions that could disrupt the global autonomous driving industry.

Social Voices: How China Is Talking About the GTR

The ADS GTR adoption was not a headline event on Chinese social media. Autonomous driving regulation, however consequential, does not generate the same viral energy as a celebrity scandal or a new smartphone launch. But within the technology and automotive communities, the reaction was substantive and telling.

On Weibo, technology commentators noted the historical significance of China moving from a standards-follower to a standards-setter. One widely-shared post from an automotive industry analyst read: "Twenty years ago, we were adapting Euro 3 emission standards. Ten years ago, we were implementing NCAP crash test protocols. Today, we are co-writing the global rulebook for autonomous driving. The speed of this transition is unprecedented." (二十年前，我们还在适应欧三排放标准。十年前，我们还在实施NCAP碰撞测试协议。今天，我们正在共同编写自动驾驶的全球规则手册。这种转变的速度是前所未有的。)

On Zhihu, a more skeptical thread questioned whether the GTR would actually benefit Chinese companies or whether it was a diplomatic gesture with limited commercial impact. The consensus in the thread, after extensive debate, was that the GTR's real value lies in reducing regulatory friction for exports. "The Chinese domestic market is large enough to sustain these companies, but to achieve global scale, they need to sell in Europe and Southeast Asia. The GTR makes that possible without rebuilding the certification stack for each country." (中国国内市场足够大，可以维持这些公司，但要实现全球规模，他们需要在欧洲和东南亚销售。GTR使这成为可能，而无需为每个国家重建认证体系。)

On XiaoHongShu (RED), travel bloggers who had experienced robotaxis in Wuhan and Guangzhou shared practical guides for foreign tourists. The posts highlighted a reality that is often lost in policy discussions: Chinese robotaxis are already a consumer experience. "I rode Apollo Go in Wuhan last week. The car arrived in 3 minutes. No driver. The seat screen showed the route in English. Cost: ¥6. This is not the future. This is now." (上周我在武汉坐了Apollo Go。车3分钟就到了。没有司机。座椅屏幕用英文显示路线。费用：6元。这不是未来。这就是现在。)

The sentiment across these platforms was consistent: pride in the technical achievement, cautious optimism about the commercial implications, and a sense that the global balance of power in automotive technology is shifting in a way that favors China.

Conclusion: The Rulebook and the Road

The ADS GTR is a dense, technical document. It contains requirements for safety management systems, performance criteria for dynamic driving tasks, validation methodologies for simulation and field testing, and audit procedures for post-deployment monitoring. It is not a manifesto. It is not a declaration of national supremacy. It is, in the most literal sense, a rulebook.

But rulebooks have power. They define what is permitted and what is not. They create the conditions under which markets operate. And they confer legitimacy on the players who can meet their requirements.

For China, the ADS GTR represents the culmination of a decade of strategic investment in autonomous driving technology, regulatory infrastructure, and international diplomacy. It is the proof that China's approach—state-directed, ecosystem-oriented, and patient—can produce outcomes that are not just domestically effective but globally influential. The fact that China co-sponsored, co-chaired, and co-authored the world's first global autonomous driving regulation is not a coincidence. It is the result of a deliberate strategy to move from the periphery of global standards to the center.

The implications extend far beyond the automotive industry. The GTR is a template for how China can engage with international technical standards in other emerging domains: artificial intelligence governance, quantum computing, biotechnology, and climate technology. The approach is consistent: build domestic capacity, demonstrate it at scale, engage in international forums as an equal partner, and align domestic standards with global frameworks so that Chinese companies face no disadvantage in foreign markets.

The road ahead is long. The GTR must be transposed into national laws. The technology must continue to mature. The business models must prove their viability. And the geopolitical tensions that shadow every aspect of US-China technology relations will not disappear simply because both countries voted for the same regulation.

But the direction is clear. On June 26, 2026, in Geneva, the world agreed on a common set of rules for autonomous driving. And China was not just in the room. China was at the table, with a pen in hand, writing the rules that the rest of the world will now follow.

For a nation that spent decades adapting to standards written elsewhere, there is no better symbol of how much the world has changed.

*Published on June 30, 2026.*

*Sources: United Nations WP.29 GRVA Session Reports (2023–2026); China Ministry of Industry and Information Technology statements; China Daily Global Edition (June 25, 2026); Yahoo Finance / Bloomberg reporting on Momenta IPO (June 2026); GlobeNewswire China Autonomous Vehicles Market Forecast (May 2026); ChinaSurvivalKit Robotaxi Guide (May 2026); Feature Asia Singapore Autonomous Shuttle Report (2025); StockTwits / Asianet News Deutsche Bank/HSBC China Auto Analysis (November 2025); UNECE GRVA Working Documents (2024–2026).*