In her December 2025 New Year address, Huawei Rotating Chairwoman Meng Wanzhou (Sabrina Meng) framed the company’s recent trajectory as a deliberate strategic reorientation. Reviewing Huawei’s resilient performance in 2025 across 5G-Advanced, HarmonyOS, intelligent driving, AI computing, and digital energy, she outlined seven priority arenas for 2026 that mark a shift from broad-based technological expansion toward focused ecosystem building and industrial enablement. This repositioning reflects both Huawei’s recalibrated corporate priorities and China’s evolving role in a fragmented global technology order shaped by intensifying U.S.–China competition.
Seen in this context, Huawei’s strategy illustrates how a leading Chinese firm leverages national strengths while navigating structural constraints. By aligning its 2026 agenda with China’s advantages in industrial scale, energy, talent, infrastructure, and open-source ecosystems—and by emphasizing deployment efficiency, system integration, and ecosystem depth—Huawei seeks to bypass weaknesses exposed by U.S. export controls and rising costs in chips and energy. The contrast with U.S. AI development bottlenecks highlights a broader competitive dynamic: Huawei’s path prioritizes diffusion, industrialization, and global accessibility, generating spillover effects that extend beyond China and reshape the terms of competition in the U.S.–China technology contest.
From Resilience to System-Building: A Summary of Meng Wanzhou’s New Year Message for 2026
In her December 2025 New Year address, Huawei Rotating Chairwoman Meng Wanzhou set out a clear and consequential reframing of the company’s strategic direction. Looking back on 2025, she emphasized Huawei’s ability to execute under sustained external pressure, while looking ahead to 2026, she articulated a deliberate shift from broad-based technological expansion toward focused ecosystem construction and industrial enablement. The message positions Huawei not merely as a technology supplier, but as a long-term architect of digital and industrial systems, reflecting both corporate recalibration and China’s changing posture in global technology competition.
Meng’s review of 2025 underscored disciplined resilience. Despite ongoing sanctions, Huawei advanced across multiple fronts: the commercial rollout of 5G-Advanced, the rapid expansion of the HarmonyOS ecosystem, the scaling of intelligent driving through Qiankun ADAS, progress in AI computing, and leadership in digital energy solutions. These achievements signal not only operational endurance, but also the maturation of Huawei’s domestic ecosystems in communications, computing, mobility, and energy—areas where scale, reliability, and integration matter more than headline-grabbing breakthroughs.
Against this backdrop, Meng outlined seven strategic arenas that will anchor Huawei’s development in 2026. These priorities—including vertical industry intelligence, the Kunpeng–Ascend open ecosystem, AI-native networks, HarmonyOS ecosystem prosperity, large-scale intelligent driving, energy-efficient AI infrastructure measured by “tokens per watt,” and liquid-cooled EV charging—reflect a decisive move away from standalone products toward systemic enablement. Collectively, they aim to embed intelligence into industry, infrastructure, and energy systems, reinforcing Huawei’s role as an ecosystem builder rather than a peripheral technology vendor.
Strategically, the address reveals a distinct model of competition. Huawei is doubling down on China’s strengths—industrial scale, engineering depth, coordinated ecosystems, and application-driven AI—while navigating structural constraints in advanced semiconductors through system-level optimization and vertical integration. This approach contrasts with U.S.-led AI trajectories that prioritize frontier models but face growing bottlenecks in energy efficiency, deployment cost, and industrial diffusion. Huawei’s emphasis on industrial AI, sovereign tech stacks, and ICT–energy–mobility integration positions it to project influence across emerging markets and shape alternative pathways for digital modernization.
Taken together, Meng Wanzhou’s New Year message marks Huawei’s transition from crisis management to proactive system-building. The evolution from capability accumulation in 2025 to strategic amplification in 2026 signals an ambition to convert resilience into lasting structural advantage. In the context of the U.S.–China technology contest, Huawei’s trajectory suggests a move beyond survival toward shaping ecosystems, standards, and industrial futures—both within China and across the global technology landscape.
Containment or Capability? The China–U.S. AI Race at a Strategic Crossroads
The competition between China and the United States in artificial intelligence has become a defining arena of great-power rivalry, often compared to a new Cold War. Gregory C. Allen argues that AI is a disruptive strategic technology akin to nuclear weapons or aerospace, and that the U.S. must slow China’s rise through strict export controls on advanced chips and semiconductor manufacturing equipment. In his view, preventing China from accessing cutting-edge AI hardware is the single most important lever the U.S. has to preserve its lead, even if this comes at economic cost. Export controls, backed by allied coordination, are framed as a containment strategy designed to block China’s path to AI and semiconductor self-sufficiency and to avoid future strategic dependence on Chinese technology.
Jensen Huang challenges this containment-first logic by reframing the China–U.S. AI race as a competition of entire national systems rather than a narrow chip blockade. He describes AI as a five-layer stack—energy, chips, infrastructure, models, and applications—and argues that long-term leadership depends on strength across all layers. From this perspective, export controls alone risk weakening U.S. firms while failing to address America’s deeper structural constraints. Huang stresses that China’s advantages in energy supply, infrastructure speed, manufacturing scale, and societal acceptance of AI give it powerful momentum that chip restrictions cannot fully neutralize.
Energy emerges as the most critical bottleneck shaping the AI race between China and the United States. While China continues to rapidly expand power generation capacity, the U.S. faces stagnant effective electricity growth and rising reliability risks, even as AI data center demand is set to surge more than twentyfold by 2030. This imbalance threatens America’s ability to scale AI deployment, regardless of its current lead in advanced models and chips. In contrast, China’s abundant and expanding electricity supply strengthens its capacity to industrialize AI across manufacturing, robotics, and other application-heavy sectors.
The debate ultimately highlights two competing strategies for winning the China–U.S. AI race: containment versus self-reinforcement. Allen prioritizes slowing China through controls and allied pressure, while Huang argues that the U.S. must win by fixing its own system—expanding energy supply, accelerating infrastructure, maintaining access to global markets, and driving AI adoption across the real economy. Recent policy shifts suggest the U.S. is moving toward a hybrid approach, balancing security restrictions with renewed emphasis on energy dominance and reindustrialization. The outcome of the AI race will likely depend less on who blocks whom, and more on which country can deploy AI at scale faster, cheaper, and across more of its economy.
Divergent Pathways: Competing Strategic Logics in AI Development
The contemporary U.S.–China technology contest is increasingly defined by two fundamentally different logics of artificial intelligence development. The U.S. approach centers on maintaining superiority at the technological frontier, reinforced by chokepoint strategies that restrict access to advanced chips, manufacturing tools, and key supply chains. While this model has preserved U.S. leadership in cutting-edge models and core research, it has also exposed structural weaknesses: rising energy constraints, slow-moving infrastructure, escalating deployment costs, and growing social and regulatory resistance to large-scale AI adoption. As a result, frontier innovation has outpaced the system-wide diffusion of AI across the broader economy.
By contrast, Huawei’s strategy—reflective of China’s wider approach—treats AI as a problem of national systems engineering rather than a single breakthrough technology. The emphasis is placed on deployment at scale, deep ecosystem coordination, and tight integration with industry, energy, and infrastructure. Temporary disadvantages at the semiconductor frontier are accepted as a strategic trade-off, offset by gains in speed, scale, resilience, and technological autonomy. This logic prioritizes making AI usable, affordable, and embedded across production systems, rather than maximizing performance at the leading edge alone.
These divergent frameworks help explain why external pressure has not derailed Huawei’s trajectory but instead reshaped it. U.S. constraints have forced a rerouting toward system optimization, vertical integration, and ecosystem depth, reinforcing a development path that is less vulnerable to chokepoints. In this sense, the competition is no longer solely about who leads in the most advanced models, but about which strategic logic can translate AI into sustained economic and industrial power.
From Chip Constraints to System Power: Huawei’s Ascend–Kunpeng Strategy
Huawei’s AI strategy illustrates how structural constraints can be transformed into sources of leverage. Confronted with restricted access to EUV lithography, advanced GPUs, and leading-edge foundry nodes, Huawei has deliberately reframed the basis of competition. Rather than pursuing absolute chip performance at the frontier, the company has shifted emphasis toward system efficiency, architectural optimization, and ecosystem control—areas where China’s scale and coordination advantages are most pronounced.
At the core of this approach are the Ascend and Kunpeng platforms, designed as open, modular, and vertically integrated computing foundations. These architectures prioritize practical efficiency—measured in tokens per watt and workload optimization—over peak theoretical FLOPs. By tightly coupling hardware, software frameworks, and industry-specific solutions, Huawei reduces dependence on the most advanced manufacturing nodes while still delivering deployable and scalable AI capability across real-world use cases.
Ecosystem expansion is the critical multiplier. Huawei has invested heavily in developer and partner mobilization, cultivating millions of developers across Ascend and Kunpeng and thousands of ecosystem partners. This broad base accelerates application development, locks in software compatibility, and embeds Huawei’s stacks deeply into enterprise and public-sector systems. Over time, the cost of switching away from these platforms rises, creating durable advantages that are independent of chip-level parity with global leaders.
In contrast, much of the U.S. AI stack remains anchored to extremely expensive GPUs, high energy costs, and manufacturing constraints that limit rapid domestic scaling. While this model excels at pushing the technological frontier, it is less optimized for cost-efficient, large-scale diffusion. Huawei’s strategy thus bypasses the GPU chokepoint not by matching it directly, but by redirecting competition toward ecosystems, deployment, and industrial integration—turning chip constraints into long-term system power within the global AI landscape.
Powering Intelligence: Energy and Infrastructure as a Strategic Divider in AI
A critical but often underappreciated dimension of AI competition lies in energy and infrastructure. China’s advantages in this domain—rapid expansion of power capacity, fast infrastructure build-out, and comparatively low-cost industrial electricity—provide a material foundation for large-scale AI deployment. These strengths allow AI to be treated not as a purely digital service, but as an integrated component of physical production systems, tightly coupled to power, transport, and industrial assets.
Huawei has positioned itself at the center of this model by directly linking AI deployment to energy and infrastructure. Through its Digital Energy division, Huawei integrates AI with power electronics, energy storage, and grid intelligence, enabling AI systems to be embedded in data centers, industrial sites, and critical infrastructure. This approach supports real-world applications such as unmanned mining operations, intelligent steel furnaces, oil and gas exploration, and energy-efficient AI data centers optimized for performance per watt rather than raw computational scale.
The strategic contrast with the United States is stark. U.S. AI growth is colliding with rapidly rising electricity demand from data centers, minimal reserve power capacity, lengthy permitting processes, and political resistance to large-scale energy expansion. These constraints raise costs and slow deployment, reinforcing an AI model that remains heavily cloud-centric and less integrated into physical industry. Even as U.S. firms lead in advanced models, their ability to industrialize AI at scale is increasingly limited by power availability.
The net effect is a divergence in AI trajectories. Huawei’s energy-integrated approach enables AI to scale in the physical economy, transforming heavy industry and infrastructure through cost-efficient, power-aware deployment. By leveraging China’s energy and infrastructure advantages, Huawei effectively bypasses a key U.S. bottleneck, turning what might appear as a secondary factor—electricity—into a decisive source of strategic leverage in global AI competition.
The Fifth Layer of AI: Manufacturing Density as Strategic Advantage
Beyond algorithms, chips, and energy, the next decisive layer of AI competition lies in real-world application density. China’s dense manufacturing base, vast deployment environments, and relatively high social acceptance of automation provide a uniquely fertile ground for embedding AI directly into production systems. This environment allows AI to be trained, tested, and refined through continuous interaction with physical processes—an advantage that cannot be replicated through simulation alone.
Huawei’s performance in 2025 offers concrete proof of this model. Its intelligent driving systems accumulated billions of kilometers of assisted driving data, while unmanned mining trucks moved from pilot projects into commercial operation. In healthcare and heavy industry, AI systems reached second-level diagnostic capability in pathology and achieved high predictive accuracy in steel furnace operations. These deployments demonstrate not isolated demonstrations, but sustained feedback loops between AI systems and complex real-world environments.
Building on this foundation, Huawei’s 2026 agenda emphasizes deepening vertical industry integration. The strategic shift is from showcasing AI capabilities to making AI a core layer of production infrastructure. By embedding intelligence across manufacturing, energy, transport, mining, and healthcare, Huawei aims to normalize AI as a routine operational tool rather than an experimental add-on, accelerating learning cycles and reinforcing ecosystem dependence.
The contrast with the United States is structural. AI deployment there is often slowed by labor politics, legal liability concerns, and public narratives that frame automation as a social threat. These frictions reduce opportunities for large-scale real-world experimentation, limiting feedback loops and anchoring much AI development in text, code, and simulated environments. The net result is a divergence in learning regimes: China trains AI on reality at scale, while the United States increasingly trains AI on representations of reality—an asymmetry that may prove decisive as AI moves from cognition into control of the physical world.
Open Source as Strategic Leverage: Huawei’s Countermove to Export Controls
Facing tightening export controls and technological exclusion, Huawei has elevated open source from a development philosophy to a strategic instrument. Its approach combines hardware openness with software open sourcing, exemplified by initiatives such as OpenEuler, Ascend toolchains, and a growing portfolio of industry-specific AI models. By open-sourcing not only platforms but also high-value assets such as pathology models and datasets, Huawei lowers barriers to entry and accelerates adoption across domestic and international markets.
The strategic logic is deliberate. Open source weakens the grip of U.S.-centric technology standards by offering credible, scalable alternatives that are not subject to unilateral restrictions. Reduced adoption costs make Huawei’s stacks attractive to developers, enterprises, and governments seeking autonomy and affordability, particularly in emerging markets. Over time, widespread use of these open platforms builds technical familiarity and operational dependence on Chinese-led ecosystems, shifting influence from proprietary dominance to ecosystem gravity.
This strategy exposes a structural contradiction in the U.S. position. Washington seeks simultaneously to constrain China’s technological rise and to preserve U.S. leadership over global standards. Yet expansive export controls incentivize third parties to reduce reliance on U.S. technologies and explore non-U.S. options. As access to American hardware, software, and tools becomes more conditional, open-source alternatives backed by Huawei appear not as second-best substitutes, but as strategic hedges against geopolitical risk.
The net effect is a reversal of intent. Sanctions designed to isolate Huawei instead accelerate the globalization of its ecosystems, turning external pressure into a catalyst for wider diffusion. By aligning open source with industrial deployment and sovereign technology needs, Huawei transforms constraints into leverage—using openness not only to bypass key weaknesses, but to reshape the competitive terrain of global technology governance.
HarmonyOS and the Exit from Platform Dependence: Huawei’s Strategy Beyond the Mobile OS
For years, China’s digital ecosystem was structurally constrained by dependence on Android and iOS, leaving core consumer and device platforms exposed to external control. Huawei’s development of HarmonyOS represents a direct response to this vulnerability, not merely as a substitute operating system, but as a foundational effort to reclaim autonomy at the platform layer. What began as a defensive necessity has evolved into a broader strategic ambition.
By 2025, HarmonyOS 5 had been deployed on more than 36 million devices, marking a transition from survival to scale. Crucially, Huawei has shifted its framing of HarmonyOS away from a “replacement OS” narrative toward an AI-native experience platform. The system is designed to operate across smartphones, vehicles, IoT devices, and industrial terminals, enabling seamless interaction among heterogeneous hardware environments. This tight coupling reflects Huawei’s emphasis on real-world deployment and system integration rather than isolated consumer applications.
Looking ahead to 2026, Huawei’s strategic bet is that HarmonyOS will function as a control layer for AI agents and a cross-device intelligence fabric. In this vision, the operating system orchestrates perception, decision-making, and execution across distributed devices, allowing AI to move fluidly between cars, factories, homes, and infrastructure. HarmonyOS thus becomes a critical enabler of embodied and industrial AI, anchoring intelligence within physical systems rather than confining it to cloud-based applications.
The contrast with U.S.-led mobile ecosystems is instructive. Android and iOS remain highly closed, geopolitically sensitive platforms, with growing risks of fragmentation outside the Western sphere. As access and governance become increasingly politicized, their global neutrality weakens. Huawei’s approach sidesteps this vulnerability by constructing an alternative stack—one rooted in openness, cross-domain integration, and AI-centric design. The net effect is not simply the creation of a new operating system, but the emergence of a parallel digital civilization stack, reducing dependency while redefining how intelligence is distributed and governed in the connected world.
Beyond Borders: The Global Spillover Effects of Huawei’s AI Model
Huawei’s AI strategy carries implications that extend well beyond China, particularly for the Global South. By combining Ascend and Kunpeng into a lower-cost, open, and politically less conditional AI stack, Huawei offers what amounts to a ready-made industrialization toolkit. This model is especially attractive to energy-rich but capital-constrained countries and late-industrializing economies that seek practical AI deployment without the financial, regulatory, and geopolitical burdens associated with Western technology ecosystems.
The contrast with the prevailing U.S.-centric AI model is structural. Nvidia-led AI stacks demand massive upfront capital investment, highly stable power grids, and compliance with Western regulatory and geopolitical frameworks. These requirements concentrate AI capability in a small number of advanced economies and hyperscale cloud providers. Huawei’s approach, by contrast, emphasizes distributed deployment, industry embedding, and infrastructure-driven scaling, allowing AI to diffuse across manufacturing, energy, transport, and public services rather than remaining centralized in data centers.
This divergence mirrors earlier waves of industrial technology diffusion, from electricity and railways to telecommunications and manufacturing automation. In each case, the dominant model was not defined solely by technical superiority, but by deployability, cost, and compatibility with local conditions. Huawei’s strategy aligns with this historical pattern, positioning AI as a general-purpose infrastructure rather than a scarce, elite resource.
As a result, standards power increasingly flows from deployment rather than decree. The actors that embed AI at scale shape workflows, interfaces, and operational defaults that others adapt to over time. Through deep industry penetration and real-world application density, Huawei’s technologies gain de facto standard-setting influence, regardless of formal international processes. This dynamic explains why Huawei’s trajectory matters globally: it is not only competing with U.S. firms, but reshaping how AI spreads, who can adopt it, and who ultimately defines the rules of the emerging intelligent economy.
Summary & Implications
Taken together, Huawei’s 2025 performance and its 2026 strategic arenas illustrate a coherent response to the structural realities of the U.S.–China technology contest. Huawei has aligned itself tightly with China’s underlying AI advantages—abundant energy, large-scale talent pools, rapid infrastructure build-out, open-source ecosystems, and a vast domestic market—while deliberately routing around exposed weaknesses such as restricted access to advanced chips, high energy costs, and regulatory or deployment frictions amplified by U.S. export controls. This alignment allows Huawei to remain competitive in AI and digital industries not by matching the United States at every technological frontier, but by optimizing for system efficiency, scale, and real-world integration.
At a strategic level, the contrast is stark: the U.S. approach seeks advantage through denial and chokepoints, while Huawei and China pursue advantage through absorption, diffusion, and deployment. Huawei’s trajectory suggests that China does not need to win the chip race outright to prevail. Instead, success hinges on dominating deployment, ecosystems, energy–AI coupling, industrial integration, and global accessibility. If this strategy holds, China’s AI industry can remain functionally independent, Huawei can sustain global relevance, and the United States will face a competitor that cannot be sanctioned out of existence, only competed against on fundamentally different terms.
References
- “Press Ahead: Spring Mountains Rise Beyond the Vast and Wild Plains, A New Year Message for 2026”. Sabrina Meng(Meng Wanzhou). December 2025. https://www.huawei.com/en/special-release/new-year-message-2026
- “Press Ahead: Spring Mountains Rise Beyond the Vast and Wild Plains, A New Year Message for 2026”. Sabrina Meng(Meng Wanzhou). December 2025. https://www.huawei.com/cn/special-release/new-year-message-2026
- “NVIDIA’s Jensen Huang on Securing American Leadership on AI”. December 3, 2025. https://www.csis.org/analysis/nvidias-jensen-huang-securing-american-leadership-ai
- “Countering China’s Challenge to American AI Leadership”, Gregory C. Allen. December 2, 2025. https://www.csis.org/analysis/countering-chinas-challenge-american-ai-leadership