The China Show on Bloomberg put a spotlight on a new reality: China’s AI pace is forcing a reset in global positioning. Zhipu’s rapid model upgrades and commercialization have turned into a market signal, while a scare trade grips parts of the US investor base that still underestimates China’s velocity. The harder read is that China’s AI stack is becoming system-level: models, chips, cloud, devices, industrial automation, and EV platforms, all reinforced by policy and scale. That is investable, and it travels.
Zhipu is now the reference point for China’s foundational model credibility, not a curiosity. The company’s GLM series has moved from demos to deployment, embedded in enterprise workflows via domestic clouds and partner ecosystems. The regulatory environment for large models is clear and accelerating approvals, enabling faster iteration and safer product rollouts in finance, healthcare, and government services. What jolts the field is not a single benchmark but the speed-to-production loop: China’s manufacturers, banks, and platforms are plugging Zhipu-like capabilities into revenue lines at scale. The outcome is compounding data advantages and a durable services business around copilots and vertical agents. That pushes the conversation from model-versus-model to platform-versus-platform. US fear trades against Chinese AI are reading headlines; Chinese operators are shipping.
Penetration is the story. Surveys show 83% of Chinese respondents already using generative AI, well above the US and global averages. That chasm matters because product telemetry and user feedback fuel faster model tuning and monetization. On the ground, Guangdong’s AI economy has grown into a multibillion-dollar cluster supporting manufacturing upgrades, while Shanghai’s robotics surge is unmistakable. In 2025, Shanghai-based AgiBot led global humanoid shipments at roughly 38% share, outpacing US entrants and proving how quickly Chinese firms commercialize hardware plus software stacks. The compute pipeline is similarly de-risking. Domestic accelerators are scaling, cloud providers are optimizing inference costs, and chipmaking capabilities continue to advance under supply chain pressure. The result is an ecosystem comfortable with constraints and skilled at extracting more performance-per-watt and performance-per-yuan across workloads.
The scare trade around China tends to manifest as bids for perceived moats in a handful of US names, plus risk premia slapped on Chinese tech overhangs. That misses two dynamics. First, export controls have acted like a forcing function in China, catalyzing local chip design, compilers, and model efficiency research that improve unit economics. Second, emerging markets are voting with deployments. From Southeast Asian fintech to Middle East logistics and African e-commerce, Chinese platforms and vendors are winning on total cost of ownership and time-to-value, often bundling AI with payments, warehousing, and devices. The capital markets read-through is clear: policy risk is real, but cash flows from AI-enabled services, EVs, and industrial automation are broadening, not narrowing.
1) Alibaba (BABA): The company’s 2025 pledge to invest $53 billion in AI, paired with a fresh Nvidia partnership, triggered a 49% surge in Hong Kong that month, the Hang Seng Tech Index’s largest gainer. Achievement: Tongyi models now underpin merchant tools, ads, and cloud copilots. Milestone: Accelerating AI attach in commerce and logistics. Global impact: Cross-border sellers on Alibaba’s platforms are using AI listings and translation to expand in Europe and the Middle East; analysts highlight a re-rating as AI monetization ramps. 2) Baidu (BIDU): ERNIE is entrenched in search, marketing, and enterprise applications while Apollo’s robotaxi stack keeps logging driverless miles in major cities. Achievement: Integrating LLMs into core adtech and maps. Milestone: New city permits for expanded autonomous operations signal regulatory confidence. Global impact: Partners across Asia leverage Baidu’s cloud AI to localize services at lower cost. 3) Tencent (0700.HK): Hunyuan powers content generation, customer service, and developer tools across WeChat, QQ, and Tencent Cloud. Achievement: Seamless AI within super-app user journeys. Milestone: Enterprise adoption of Hunyuan in manufacturing and retail scenarios. Global impact: Gaming studios deploy Tencent’s AI pipelines for asset creation, compressing dev cycles and export capacity. 4) Xiaomi (1810.HK): The company fuses on-device AI with an expanding EV strategy. Achievement: AI-accelerated imaging and assistants in smartphones, plus autonomous features in the SU7 line. Milestone: Scaled first-year EV deliveries establish Xiaomi as a credible software-defined vehicle player. Global impact: AI-forward devices in Europe and MENA build a bridge from consumer to mobility ecosystems. 5) SMIC (0981.HK): China’s leading foundry is advancing process capabilities amid intense demand for domestic compute. Achievement: Progress on mature and more advanced nodes supports a wide range of AI and edge chips. Milestone: Increased orders tied to AI inference and IoT. Global impact: Supply chain diversification for global OEMs hedging geopolitical risk. 6) Cambricon (688256.SS): A key domestic AI chip designer focused on accelerators for training and inference. Achievement: New-generation MLUs targeting better efficiency per dollar in data centers and at the edge. Milestone: Design wins with top internet platforms and state clouds. Global impact: Options for emerging markets to deploy AI without premium Western pricing. 7) iFlytek (002230.SZ): A national champion in speech and education AI with its Spark model series. Achievement: Multimodal upgrades spanning voice, code, and tutoring services. Milestone: Broad adoption by schools and hospitals seeking localized, safe AI. Global impact: Language tech that travels across ASEAN and Africa, boosting digital inclusion. 8) BYD (1211.HK): The world leader in new energy vehicle units is embedding AI across manufacturing, energy management, and ADAS. Achievement: Vertical integration lowers costs for AI-heavy features. Milestone: Overseas plants and distribution deepen. Global impact: Affordable EVs with smart features raise adoption in Latin America, Southeast Asia, and Europe’s value segments.
Capital is already adjusting. Mainland investors are leaning into AI supply-chain proxies on STAR and ChiNext, while global funds are rebuilding China tech allocations selectively through Hong Kong. The rationale is straightforward: AI is no longer a single vertical; it is a capability layer touching payments, logistics, content, industrial controls, and autos. That breadth lowers portfolio idiosyncratic risk even if any one line item wobbles. For allocators, the investable expression is a barbell: platforms with user-scale data on one side and hard-tech enablers on the other. Volatility from headlines is part of the price of admission, but the earnings power sits with companies turning AI into productivity and transaction growth.
Beijing’s policy mix is extending the runway. Compute buildouts, data governance frameworks, and incentives for smart manufacturing are pulling forward demand while setting standards that favor scaled players. That creates a platform premium for firms that can comply, integrate, and export. The link between AI and real-economy outcomes is explicit: better factory yields, faster customs clearance, smarter energy dispatch. This is not hype; it is operating metrics. With financing costs stabilizing and capital markets reform ongoing, the valuation gap between China’s AI leaders and global peers remains an opportunity, not a verdict.
Chinese AI is not confined to the mainland. Southeast Asian retailers are using Chinese recommendation engines to lift conversion. Middle Eastern utilities are adopting Chinese grid AI to smooth demand spikes. African fintechs are launching with Chinese risk models packaged through cloud APIs. Robotics makers in Shanghai are selling into European warehouses on cycle time and reliability. These are sticky deployments that underwrite multi-year revenue. Zhipu’s rise underscores what the Bloomberg segment captured: the race is now about who can integrate models into mission-critical workflows at the lowest cost and highest pace. On that measure, China’s innovation engine is running hot, its scale is compounding, and its leadership is increasingly priced in global outcomes rather than press releases.
