AI and Robotics in China: How Beijing Is Turning Automation Into National Power

China installed 295,000 industrial robots in 2024. The United States installed 34,200. That is not a technology gap — it is not even primarily an investment gap. It is the visible output of a twenty-year policy architecture that has systematically converted national strategic intent into physical manufacturing infrastructure, and it is accelerating. In 2026, China's 15th Five-Year Plan — formally adopted in March and placing "embodied intelligence" among its top ten prioritised industry tracks — has converted what was a manufacturing lead into a formal national policy architecture, making AI-powered robotics not a sector China is developing but the structural framework through which it intends to develop everything else.

The most consequential thing China is doing in AI robotics is not the robot half-marathon in Beijing or the armed quadrupeds at the World Defense Show. It is the Humanoid Robot and Embodied Intelligence Standardization Technical Committee established in December 2025, now leading formulation of IEC global standards for elder-care robots. The country that writes the standards for how robots must behave, interoperate, and share data has leverage over every company that wants to sell robots anywhere in the world — leverage that persists long after any hardware advantage has been competed away. Understanding China's robotics position requires separating three things that coverage routinely conflates: the industrial deployment reality, the policy architecture enabling it, and the military translation that is simultaneously more documented and more constrained than most Western analysis acknowledges.

What Is China's AI Robotics Strategy?

China's AI robotics strategy is a state-coordinated programme that uses the systematic deployment of industrial robots and AI-powered automation as the primary mechanism for solving three structural national problems simultaneously: a shrinking manufacturing workforce driven by demographic decline, a persistent dependence on foreign technology in strategically sensitive sectors, and a military modernisation requirement that cannot be met through conventional hardware procurement alone. It exists because each of those three pressures has reached a threshold at which incremental policy responses are inadequate — and because the combination of China's manufacturing scale, domestic supply chain depth, and centralised policy coordination creates conditions in which robotics deployment can be accelerated at a rate that no decentralised market economy can replicate. For any business whose supply chain touches Chinese manufacturing, any investor with exposure to global robotics, or any policymaker assessing the competitive implications of China's industrial trajectory, this strategy is not a future scenario — it is the operating environment.

What 295,000 Robots a Year Actually Does to a Competitive Position

The number that reframes the conversation is not the annual installation figure — it is the stock. China's operational stock of industrial robots surpassed 2 million units in 2024, approximately 4.5 times more than Japan, the world's second-largest robot stock, according to the International Federation of Robotics World Robotics 2025 Report. Every robot on a factory floor is generating operational data. That data trains the AI systems that make the next generation of robots more capable and easier to deploy. The flywheel is not theoretical — it is running, and it runs faster the larger the installation base becomes.

How many robots does China have? China's operational stock of industrial robots surpassed 2 million units in 2024 — approximately 4.5 times more than Japan, the world's second-largest robot stock — and China installed 295,000 new industrial robots in 2024, representing 54% of all global deployments that year, according to the IFR World Robotics 2025 Report. Chinese manufacturers now supply 57% of domestic robot sales, up from 30% in 2020.

The supplier shift inside that number matters as much as the volume. Chinese manufacturers' domestic market share in robot sales climbed from 30% in 2020 to 57% in 2024. In the metal and machinery sector, Chinese robot suppliers now hold 85% domestic market share — territory that was held by ABB, KUKA, and Fanuc a decade ago. This is not import substitution as a trade policy. It is the creation of a domestic robotics supply chain that makes future deployment cheaper, faster, and independent of foreign supplier relationships that could be disrupted by sanctions, export controls, or geopolitical pressure. The comparison with what happened in solar panels and 5G — where China's combination of domestic scale, state support, and supply chain integration produced cost structures that foreign competitors could not match — is not a conspiracy theory. It is a pattern with documented precedents.

The contrast with the rest of the world is not uniform, which is the detail that makes the analysis useful rather than merely alarming. Western European robot density reached a record 267 robots per 10,000 manufacturing employees in 2024, ahead of China's approximately 470 — though China's density is growing faster, according to the IFR World Robotics 2025 April 2026 update. European industrial robot installations fell 8% to 85,000 units in 2024, while China's first-half 2025 industrial robot production surged 35.6% year-on-year to nearly 370,000 units, according to China's National Bureau of Statistics. The US installed 34,200 units in 2024 at a robot density of approximately 295 per 10,000 workers. China installed more industrial robots in 2024 than the United States has installed in the past decade combined — and that gap is accelerating, not closing.

What the IFR's Takayuki Ito said publicly — that the data shows "a growing automation gap between China and other industrial economies" — understates what the compound trajectory implies for the 2030s. The gap between deployment rates determines future capability gaps, because real-world deployment is the primary source of the training data that improves the AI systems running on those robots.

How the 15th Five-Year Plan Functions as a Procurement Directive, Not an Aspiration

The 15th Five-Year Plan, adopted in March 2026, contains language on embodied intelligence that the Diplomat's March 2026 analysis correctly described as "not aspirational language — it is a procurement directive." Box 3, Item 02 of the plan specifies that China will "coordinate the layout of embodied intelligence training grounds, promote virtual-real fusion collaborative training and evolution, develop integrated big-brain/small-brain embodied models and algorithms, tackle key technologies in the body and core components, and accelerate the upgrade and deployment of humanoid robots and other form-factor products." That level of specificity in a Five-Year Plan means subordinate sectoral plans at the ministry level and provincial plans at the regional level are now mandated to align. IFR President Takayuki Ito confirmed this: "thousands of subordinate sectoral and regional plans are now mandated to align with the objectives."

The provincial competition this generates is the mechanism most observers miss. Guangdong has signalled embodied AI as a priority and launched a 10 billion RMB AI and Robotics Industry Fund in Shenzhen in early 2025. Beijing's Yizhuang development zone is targeting designation as "a global first-class embodied intelligent robotics industrial new city." Over 150 humanoid robot companies now operate in China, supported by more than 40 state-funded robot training centres generating millions of real-world training data entries. The state-backed fund committed to humanoid robots, industrial automation, and embodied AI is estimated at 1 trillion yuan — approximately $138 billion — dwarfing any comparable Western initiative. In the Chinese system, provincial competition to implement central targets creates implementation dynamics that central planning alone cannot generate: local governments have career incentives to exceed national targets, not merely meet them.

The standards campaign is the element of this strategy that receives the least coverage and carries the most durable strategic significance. China's Ministry of Industry and Information Technology established a dedicated Humanoid Robot and Embodied Intelligence Standardization Technical Committee in December 2025 and released the first national standard system covering the humanoid robot industry's entire lifecycle by March 2026. China is now leading formulation of IEC global standards for elder-care robots and actively shaping international norms for robot safety, interoperability, and data governance. This is standard power — the form of geopolitical leverage that accrues to the country that sets the technical standards around which a global industry must organise, and that persists after any hardware advantage has been competed away. The pattern replicates China's successful campaigns in 5G and high-speed rail: establish the domestic standard first, build commercial scale around it, then export it as the de facto international norm. When a Vietnamese care facility buys elder-care robots in 2030, the interoperability standards those robots must meet will have been written in Beijing in 2026.

The Military Dimension: What the Procurement Records Actually Show

The Reuters investigation published in October 2025 — reviewing hundreds of research papers, patents, and procurement records from the PLA Procurement Network — provides the most credible public window into how China is integrating AI into military systems. What it shows is systematic, documented, and specific. Two dozen tenders and patents demonstrate the PLA's effort to integrate AI into drones for target recognition, tracking, and formation flying with minimal human intervention. In November 2024, the PLA issued a tender for AI-powered robot dogs to scout in packs and clear explosive hazards. In February 2025, state-owned defence giant Norinco showcased the P60 military vehicle capable of autonomous combat-support operations at speeds up to 50 kilometres per hour, powered by DeepSeek AI running on Huawei chips. At the World Defense Show 2026 in Riyadh, a state-linked manufacturer introduced the PF-070 — a four-legged combat robot armed with four anti-tank guided missiles, marketed as a production-ready platform for international sale.

The civil-military fusion architecture that enables this is structural, not opportunistic. Chinese LiDAR companies — whose products directly enable robot navigation and targeting — account for nearly 80% of global sales, per Foundation for Defense of Democracies analysis from May 2026, giving the PLA a deep, price-competitive sensor base independent of foreign supply. The PLA's Integrated Command Platform and the Information Support Force established in 2024 are building the digital command-and-control layer within which autonomous systems operate. Beihang University's 2025 patent filing demonstrating DeepSeek-powered drone swarm decision-making for targeting "low, slow, small" threats is not a university research project in any conventional sense — Beihang is one of China's "Seven Sons," a group of universities under formal partial PLA supervision.

What the coverage tends to overstate is the direct operational capability translation. Chinese defence leaders have publicly committed to maintaining human control over weapons systems — a commitment that the Reuters review notes is under genuine tension as AI systems become more capable, but that remains the stated doctrine. More concretely: autonomous drone swarms that can recognise and track targets in formation are not the same as autonomous drone swarms that can prosecute targets with military effectiveness in contested electronic warfare environments. The gap between the capability demonstrated in procurement tenders and the capability required for operational use in a high-end conflict is real, even if that gap is closing faster than Western militaries have planned for.

Where China's Own Government Acknowledges the Limits — and Why That Doesn't Make It Less Strategic

The most credible pushback against the "running away" framing comes not from Western sceptics but from the IFR's own analysis of China's 15th Five-Year Plan — analysis conducted with access to the same Chinese data that generates the alarming headlines.

The dominant narrative in Western defence and technology commentary holds that China is, in the words of Andreessen Horowitz's Martin Casado and Anne Neuberger, "running away with the hard power part of AI — robotics." That framing is accurate about the industrial deployment gap but misleading about the mechanism and the timeline. The IFR explicitly states: "mass adoption as universal humanoid factory helpers or in private households will not happen within the near- and medium-term future. The 15th Five-Year plan sees the commercialization of humanoid robots rather towards the end of the plan's period." China's own government, in the Five-Year Plan itself, targets humanoid robot commercialisation "towards the end of the plan's period" — meaning 2029 to 2030 at the earliest. The humanoid robot half-marathon in Beijing and the viral demonstrations of robots dancing at the Lunar New Year celebration are exactly what the IFR identifies them as: demonstrations of competence and global competitiveness, not operational readiness.

The genuine limits are the same ones that constrain every robotics ecosystem: the embodiment gap between what AI can compute and what a robot can reliably execute in unstructured physical environments, the hardware constraints in tactile sensing and dexterity, and the common sense problem that no training dataset fully solves. In China's case, these limits are partially offset by the sheer volume of real-world deployment data being generated by 2 million operational industrial robots — but that data advantage applies to structured manufacturing environments, not to the general-purpose physical intelligence that the humanoid demonstrations suggest.

The structural argument that the "running away" framing misses is subtler and more consequential. China does not need humanoid robots to be fully operationally capable to capture the strategic benefits of leading the field. It needs to be far enough ahead that it sets the standards, trains the engineers, builds the supply chains, and establishes the data assets that define what "advanced" means as the technology matures. The country that leads the first generation of a technology typically sets the terms for the second and third generations — not because it maintains a technical lead, but because the infrastructure, talent, and institutional knowledge it accumulates are sticky in ways that point-in-time capability comparisons don't capture.

There is a version of this argument where China's robotics position peaks too early — where the current generation of industrial robots becomes commodity hardware, where humanoid deployment timelines slip to 2035, and where Western companies with different institutional incentives produce the next architectural breakthrough. That version is possible. It is also exactly what people said about China's solar panel strategy in 2015.

What China's scale, policy architecture, and deliberate standards campaign actually mean for any non-Chinese business operating in sectors that will be touched by robotics — which is most sectors within a decade — connects directly to the questions this site has been tracking across deployment, data, and the strategic implications of physical AI at scale.

This development reinforces:

• Who Owns Robot Data: China's 2 million operational industrial robots are generating the largest single-country operational robot data asset in the world, and who owns and can access that data — under China's data governance framework — is the question that determines who controls the AI systems trained on it.
• Is the AI Robotics Investment Boom a Bubble?: China's whole-of-nation mobilisation — including a 1-trillion-yuan state fund and provincial competition dynamics — represents exactly the kind of infrastructure residue that survives speculative cycles; understanding China's robotics position changes how to evaluate which Western investments are competing against state-backed scale and which are not.
• Robots in Ukraine War Effectiveness: The documented PLA integration of AI into drone swarms and autonomous ground vehicles directly extends the lessons from Ukraine — where autonomous systems at scale have already demonstrated operational effectiveness — into the specific context of China's military modernisation and the capability translation question this article identifies as underexamined.

The procurement record that Reuters reviewed in October 2025 showed something specific: the PLA issued a tender for AI-powered robot dogs that would "scout in packs." That image — a pack of autonomous scouts, coordinated by AI, operating at the edge of human oversight — is the one that headlines use to make China's robotics strategy frightening. It is real. It is also not the most consequential thing in the document. The most consequential thing is that the same country issuing that tender is simultaneously writing the international standards that will determine how every elder-care robot, logistics robot, and manufacturing robot in the world must behave, communicate, and share data by 2030. One of those developments is easier to make into a defence budget argument. The other is the one that shapes the terms of competition for the next twenty years.

1. How many robots does China have? China's operational stock of industrial robots surpassed 2 million units in 2024 — approximately 4.5 times more than Japan, the world's second-largest robot stock — and China installed 295,000 new industrial robots in 2024, representing 54% of all global deployments that year, according to the International Federation of Robotics World Robotics 2025 Report. Chinese manufacturers now supply 57% of domestic robot sales, up from 30% in 2020. China's first-half 2025 industrial robot production surged 35.6% year-on-year to nearly 370,000 units, according to China's National Bureau of Statistics.

2. What is China's AI robotics strategy? China's AI robotics strategy uses systematic deployment of industrial robots and AI-powered automation as the primary mechanism for addressing three structural national problems simultaneously: a shrinking manufacturing workforce, dependence on foreign technology in strategic sectors, and military modernisation requirements. The 15th Five-Year Plan (2026–2030), adopted in March 2026, formally elevates "embodied intelligence" — AI-powered physical robotics — to one of China's top ten prioritised industry tracks and mandates alignment of thousands of subordinate sectoral and provincial plans. The strategy's most underappreciated dimension is standard power: China is now leading formulation of IEC global standards for elder-care robots and robot interoperability, following the same playbook it used in 5G and high-speed rail.

3. Is China ahead in military robotics? China is ahead in integration of AI into specific military platforms — particularly drone swarms and autonomous ground vehicles — but the direct translation to operational military capability is more conditional than headlines suggest. A Reuters review published in October 2025 of hundreds of PLA procurement records and patents documented systematic integration of DeepSeek AI into drone targeting, robot dog scouting, and autonomous formation flying, while noting that Chinese defence leaders maintain a stated commitment to human control over weapons systems. The gap between capability demonstrated in tenders and capability operational in high-end conflict environments is real, though it is closing faster than Western military planning has assumed.

4. What is China's 15th Five-Year Plan for robotics? China's 15th Five-Year Plan (2026–2030), adopted by the National People's Congress in March 2026, places "embodied intelligence" — AI-powered physical robotics — among its top ten prioritised new industry tracks. The plan is not aspirational: it is a formal procurement directive that mandates thousands of subordinate sectoral and provincial plans to align with robotics targets, supported by a state-backed fund estimated at 1 trillion yuan ($138 billion). The IFR's Takayuki Ito confirmed that under the new framework, "China is shifting its focus from traditional industrial automation to high-end, intelligent robotics integrated with artificial intelligence." The plan targets humanoid robot commercialisation towards the end of the plan period — 2029 to 2030 — rather than immediate mass deployment.

5. How does China's robot count compare to the US? China installed 295,000 industrial robots in 2024, while the United States installed 34,200 — roughly one-tenth of China's annual rate — according to the International Federation of Robotics World Robotics 2025 Report. China's robot density stands at approximately 470 robots per 10,000 manufacturing workers, compared to approximately 295 in the United States. China has approximately 4.5 times more operational industrial robots than Japan and roughly 5 times more than the US. This gap is widening: China's industrial robot production surged 35.6% in the first half of 2025 alone, while US installations declined 9% in 2024.

6. What is standard power in robotics? Standard power is the strategic leverage that accrues to a country or company that sets the technical standards — for safety, interoperability, and data governance — around which a global industry must organise. China established a Humanoid Robot and Embodied Intelligence Standardization Technical Committee in December 2025 and by March 2026 had released the first national standard system covering the humanoid robot industry's entire lifecycle, while leading formulation of IEC global standards for elder-care robots. Standard power persists after hardware advantages are competed away: any company that wants to sell robots globally must meet the standards that China wrote domestically, giving Beijing ongoing influence over the terms of global robotics competition independent of who builds the best individual machine.