China’s EV industry delivered a packed set of signals on June 3-4, spanning robotics, autonomous driving, AI funding, and fast-moving vehicle demand. BYD said it is developing humanoid robots and could one day sell them through its dealer network, while XPeng used fresh European survey data and a factory livestream to argue that global L3 autonomous driving could begin rolling out in Q1 2027. At the same time, GAC posted a 71.64% year-on-year surge in May new-energy vehicle sales, DeepSeek reportedly launched a RMB 50 billion fundraising round, and Chinese autonomous driving suppliers such as Momenta and DeepRoute.ai continued expanding at home and abroad.
BYD’s Next Bet: From EVs to Humanoid Robots
BYD is no longer talking only about batteries, plug-in hybrids, and EV platforms. According to remarks from BYD executive vice president Stella Li, the company is actively developing humanoid robots, viewing the field as a natural extension of capabilities it has already built in automotive AI, software-hardware integration, and large-scale manufacturing.
That matters because BYD’s edge is not just concept development, but industrialization. Li argued that success in robotics will depend on three core pillars:
- Manufacturing capability
- Software and hardware integration
- AI-related vehicle technology that can transfer into robotics
Perhaps the most striking detail was the go-to-market idea: if BYD believes humanoid robots are ready for home use, it could sell them through its existing dealer network. That suggests BYD sees robotics not as a lab experiment, but as a potentially commercial consumer product.
Li also said BYD could build an open platform, supporting both in-house robot development and partnerships with outside companies. In other words, BYD may be positioning itself less like a single-product maker and more like a robotics ecosystem player.
Why BYD’s robotics move is credible
BYD is better placed than many startups because it already controls major parts of the EV value chain:
- Battery manufacturing
- Power electronics
- Electric motors
- Vehicle software integration
- High-volume assembly operations
In practical terms, humanoid robots need many of the same disciplines as modern electric vehicles: sensing, actuation, energy management, embedded computing, and cost-down manufacturing.
XPeng: L3 Autonomy Is Coming, but Europe Remains Skeptical
XPeng paired product optimism with a reality check on consumer sentiment. The company released a survey covering more than 5,000 respondents across six European countries, showing that awareness of AI is high, but trust in fully AI-controlled driving remains low.
Key findings from XPeng’s Europe survey
- 82% of respondents are familiar with AI
- Only 13% are willing to try a fully AI-controlled autonomous car
- In a comparable China survey, that figure was 70%
- 53% expressed little or no trust in AI taking over driving
- 53% said they accept driver-assistance features such as adaptive cruise control and lane-keeping
- 61% cited “loss of human control” as their biggest concern
The differences within Europe are also notable. Spain appeared relatively open, with 63% willing to ride in an autonomous car, while the UK was much more cautious at 34%.
Europe autonomous-driving sentiment at a glance
| Metric | Europe | China comparable survey |
|---|---|---|
| Familiar with AI | 82% | N/A |
| Willing to try fully AI-controlled driving | 13% | 70% |
| Little or no trust in AI taking over driving | 53% | N/A |
| Accept ADAS features like ACC/lane keeping | 53% | N/A |
| Main concern: loss of human control | 61% | N/A |
XPeng president Brian Gu’s conclusion was telling: Europe does not necessarily need more features first; it needs more exposure, more transparency, and stronger reassurance that the driver remains in control.
That aligns with comments from chairman and CEO He Xiaopeng, who said global implementation of L3 autonomous driving is expected to begin in Q1 2027. He also disclosed that XPeng’s second-generation VLA model, a vision-language-action foundation model, is already being tested overseas.
The broader takeaway is that China’s EV makers may be technologically ready to push advanced driver assistance and conditional automation faster than overseas consumers are psychologically ready to accept it.
XPeng GX Demand Is Strong, but Delivery Discipline Matters
Separate reporting from XPeng’s Guangzhou factory livestream highlighted another challenge familiar to fast-growing EV brands: balancing demand, production, and customer fairness.
After strong orders for the XPeng GX, online rumors claimed buyers could pay an extra RMB 20,000 to jump the delivery queue. He Xiaopeng publicly denied this, saying XPeng strictly prohibits any form of paid priority delivery and allocates production according to order time.
He said even personal contacts had asked to pay extra for faster delivery, but those requests were rejected. His concern was straightforward: once an automaker allows queue-jumping on a hot model, it risks normalizing markups across popular trims and destabilizing the market.
What XPeng said about GX delivery
- No paid queue-jumping is allowed
- Delivery sequence follows order timing
- Current delivery wait has reached 35 weeks
- Production capacity is under pressure
- XPeng has opened new tooling and production lines for key parts
- Delivery speed should improve moderately in the next 2-4 weeks
- The GX reportedly includes 200 more vehicle quality inspection items than the industry average
- Overseas blind orders have exceeded 1,000 units
For a Chinese EV brand trying to build global credibility, that message matters. Refusing dealer-style markups and preserving transparent delivery rules supports brand trust, especially as Chinese automakers expand into Europe and other export markets.
GAC’s Sales Show the Strength of China’s NEV Shift
While autonomy and robotics grabbed headlines, core market momentum remains essential. GAC Group reported that May 2026 vehicle sales rose 8.18% year on year to 127,330 units, while cumulative sales for the year reached 628,219 units, up 3.80%.
The standout figure was in new-energy vehicles: GAC sold 46,056 NEVs in May, up 71.64% year on year.
GAC May 2026 performance
| Metric | Result | YoY change |
|---|---|---|
| Vehicle production | 122,675 | -4.59% |
| Cumulative production | 651,372 | +0.46% |
| Vehicle sales | 127,330 | +8.18% |
| Cumulative sales | 628,219 | +3.80% |
| NEV sales | 46,056 | +71.64% |
Brand-level performance was mixed:
- GAC Honda declined year on year
- GAC Toyota and GAC Trumpchi were roughly flat
- GAC Aion surged 74.76% year on year
This is a familiar pattern in China’s auto market. Joint-venture legacy brands are increasingly under pressure, while domestic EV-focused marques are driving growth.
Autonomous Driving Suppliers Keep Scaling Fast
China’s smart-EV race is no longer defined only by carmakers. The supplier ecosystem is becoming just as important.
Seres-ByteDance brand may use DeepRoute.ai
According to Chinese media reports, a new vehicle brand being developed by Seres and ByteDance is expected to use intelligent driving technology from DeepRoute.ai. The company already has mass-production partnerships with Great Wall Motor, Leapmotor, and smart, and says its urban navigation-assisted driving system has reached more than 10 vehicle models with 300,000 vehicles on the road.
DeepRoute.ai, founded in 2019 and based in Shenzhen, has completed eight financing rounds totaling more than USD 1 billion, including a USD 300 million round led strategically by Alibaba in 2021.
Momenta broadens test permits and global rollout
Momenta said it has received road-testing and demonstration approval for intelligent connected vehicles in Wuxi, Jiangsu province, and has already begun local testing. The company plans to launch advanced autonomous driving in multiple new cities and regions globally during 2026.
Its scale metrics are increasingly difficult to ignore:
- More than 800,000 production intelligent-driving systems deployed
- More than 70 mass-produced vehicle models delivered
- More than 200 vehicle programs awarded
- Commercial presence across more than 10 countries and regions
Momenta also said its R7 reinforcement-learning world model, released in the first half of this year, has been applied to L4 autonomous driving practice. The company says the model improves understanding of physical properties, causal motion, and potential interactions, helping with prediction, planning, lane selection, and tight-space maneuvering.
Its partner list is equally significant, including:
- Uber
- Grab
- Lumo
- SAIC’s mobility platform Xiangdao Chuxing
- Mercedes-Benz
The implication is clear: Chinese autonomous driving companies are evolving from domestic ADAS suppliers into global mobility technology platforms.
AI Capital Is Flowing Into the EV Ecosystem
The EV story in China is now deeply intertwined with AI infrastructure. Reuters and other outlets reported that AI startup DeepSeek has launched its first external fundraising round, targeting roughly RMB 50 billion, or about USD 7.4 billion.
If completed at the reported terms, DeepSeek’s valuation could reach RMB 350-400 billion, potentially setting a record for a single domestic AI financing round.
Reported DeepSeek financing structure
| Investor/Item | Reported amount |
|---|---|
| Total fundraising target | RMB 50 billion |
| Founder Liang Wenfeng commitment | RMB 20 billion |
| Tencent potential investment | RMB 10 billion |
| CATL planned investment | RMB 5 billion |
| Implied valuation after round | RMB 350-400 billion |
CATL’s involvement is particularly relevant for EV watchers. The battery giant is increasingly expanding into AI data-center power supply and energy storage, aiming to build a tighter link between electricity infrastructure and computing demand. That “power + compute” strategy could become a major industrial theme as China scales both EV manufacturing and AI workloads.
ByteDance is pushing in a similar direction. Its Volcano Engine MaaS business reportedly has a 2026 revenue target of RMB 15 billion, up sharply from about RMB 1.5 billion in 2025. Seedance 2.0, its video-generation model, is said to be generating more than RMB 1 billion in monthly revenue, with further growth still underway.
For automakers and suppliers, this matters because the competition for AI talent, computing resources, and model capabilities increasingly overlaps with the competition for smarter cockpits, better autonomous-driving stacks, and robotics platforms.
BYD’s Reputation Management Shows Another Side of Competition
One smaller but revealing item involved BYD’s legal campaign against online defamation. Chinese reports said an auto blogger publicly apologized to BYD for publishing unverified and damaging content in 2023-2024 and posted more than ten apology videos.
The case fits a broader pattern. Several bloggers have reportedly faced lawsuits and sizable compensation judgments for publishing false or insulting material about the company. Reported compensation figures in other cases included roughly RMB 2.0187 million, RMB 2.01 million, RMB 2 million, and RMB 313,800.
This is worth noting because brand warfare in China’s EV market is no longer limited to pricing and product launches. It now extends to social media narratives, legal enforcement, and reputation defense.
Why This Matters Globally
Several themes from these stories have implications far beyond China.
1. Chinese EV companies are expanding into adjacent technologies
BYD’s humanoid robot project shows that leading Chinese automakers are no longer just car manufacturers. They are becoming broader intelligent hardware and AI companies.
2. China may lead on deployment before the West leads on acceptance
XPeng’s survey suggests European consumers remain cautious about autonomous driving even as Chinese companies move toward L3 deployment timelines. The bottleneck may be trust, regulation, and user education rather than core technology alone.
3. Scale is reinforcing China’s competitive edge
GAC’s NEV growth, XPeng’s order volume, and Momenta’s deployment numbers all point to one advantage that is hard to replicate elsewhere: China’s ability to test, iterate, and industrialize at enormous scale.
4. AI, batteries, and mobility are converging
DeepSeek, CATL, ByteDance, and automakers are increasingly operating in overlapping strategic territory. The future EV leader may not be the company with the best hardware alone, but the one that best combines energy systems, AI models, software integration, and manufacturing execution.
What to Watch Next
The next few quarters should reveal whether these announcements become durable trends.
Key things to monitor include:
- Whether BYD formalizes a robotics roadmap or partnerships
- How quickly XPeng can reduce GX delivery times and convert overseas interest into real sales
- Whether L3 regulations and liability frameworks start aligning globally for a 2027 rollout
- How fast Momenta and DeepRoute.ai can translate technology wins into wider international commercial deployments
- Whether AI capital inflows reshape the competitive balance in smart EV software and robotics
China’s EV industry is no longer just about selling more electric cars. It is becoming a contest over who can build the most scalable intelligent-machine platform, from batteries and vehicles to autonomous systems, cloud AI, and potentially even humanoid robots.
