The Second World Humanoid Robot Games concluded on August 26, bringing another high-profile moment for humanoid robotics and embodied intelligence. On the track, robots in sprinting, high jump, long jump and other events repeatedly broke human world records. Beyond the arena, a more consequential transition is taking shape. Humanoid robots and embodied intelligence systems are moving from demonstrations into duty, taking roles as component assemblers on factory lines, sorters in warehouse logistics, rehabilitation assistants in healthcare, and caregiving helpers in elderly support. That shift has opened a parallel test: whether these machines can be safe, reliable, scalable and commercially sustainable.
Insurance companies are stepping into that test as enablers. They are building protection networks for humanoid robots and embodied intelligence, aiming to remove obstacles that stand between laboratory breakthroughs and industrial deployment. The coverage described by industry participants is not limited to the application side. It extends across the full chain of research, development and production for humanoid robots and embodied intelligence. Policies include property damage insurance and third-party liability insurance designed for risks such as hardware destruction, electrical failure and cybersecurity incidents. They also reach into earlier stages of innovation, including research and pilot testing.
The development matters because embodied intelligence is not simply another piece of machinery. A humanoid robot combines hardware, software, algorithms, sensors, communication systems and human interaction. Its value comes from mobility, autonomy and adaptability, but those same qualities create unfamiliar risk profiles. Insurance is therefore becoming part of the commercial infrastructure for embodied intelligence, not merely a back-end financial product. It helps companies transfer risk, reassure customers, protect assets and build confidence as robots enter real workplaces.
1. From Race Track to Workplace: Embodied Intelligence Enters a New Test
The Second World Humanoid Robot Games offered a visible stage for embodied intelligence. Robots competed in events that required balance, propulsion, control and rapid decision-making. The results drew attention to how quickly humanoid platforms are improving. Yet the more important story may be what happens after the medals are awarded. The same technologies that can sprint, jump and navigate an arena are also being tested in factories, warehouses, hospitals and care settings.
In factories, humanoid robots can act as component assemblers. In warehouse logistics, they can work as sorters. In healthcare, they can support rehabilitation. In elderly care, they can serve as helpers. Each of these environments introduces different risks. A factory may involve heavy equipment, repetitive motion and electrical systems. A warehouse may involve moving vehicles, crowded aisles and shifting inventory. A healthcare setting may involve vulnerable people and strict safety expectations. An elderly care setting may involve unpredictable human behavior and emotional sensitivity.
This is why the arrival of embodied intelligence in daily operations is accompanied by a safety and reliability test. The market must determine not only whether robots can perform tasks, but also whether they can do so consistently, affordably and with acceptable risk. Insurance enters that equation because it can absorb some of the financial consequences of failure, accident or interruption. For companies deploying embodied intelligence, insurance can make the difference between a pilot project and a scalable business.
The insurance industry is responding with products that cover the application side and the industrial chain. On the application side, insurers are offering protection against hardware damage, electrical faults and cybersecurity risks. They are also providing third-party liability coverage. On the industrial chain side, coverage is extending into research, development, production and testing. This full-chain approach reflects a recognition that embodied intelligence risks do not begin only when a robot is switched on in a customer’s facility. They begin much earlier, in laboratories, test beds and production lines.
2. Insurance Becomes a Strategic Partner of the Games
As a strategic partner of the Second World Humanoid Robot Games, PICC considered the unique risk characteristics of a humanoid robot competition and customized a comprehensive risk protection plan. The insurer drew on experience from the first event and designed coverage that included embodied intelligence robot property damage insurance, embodied intelligence robot third-party liability insurance, and artificial intelligence service liability insurance. Together, these products formed a multi-dimensional protection system covering property loss, third-party liability, personal accidental injury and cybersecurity risk.
A PICC P&C executive said the company used its integrated financial strengths to build an event protection system across three dimensions. In underwriting, the insurer created a combined insurance plan for both competition time and daily operations. In claims, it established a green channel with rapid response mechanisms. In services, it provided full-cycle and efficient professional technical support to ensure seamless event operations.
The on-site service team remained present throughout the event. It followed the progress of each competition, monitored developments, provided risk reminders, handled consultation and coordination, and conducted on-site surveys. If an unexpected situation occurred, such as a person being accidentally injured or robot equipment being damaged, the team responded immediately. It simplified coordination processes, completed on-site surveys and verification quickly, and helped protect both people and equipment. The goal was to remove worries for participating teams so they could focus on technical competition.
The event therefore became a live demonstration of insurance supporting technology. It showed how risk management can be integrated into a fast-moving, highly technical environment. It also offered a preview of how insurers may support embodied intelligence after the competition ends. The arena is a controlled setting, but the knowledge gained there can inform coverage for factories, warehouses, hospitals and homes, where conditions are less predictable and the stakes are higher.
3. Market Potential Drives Full-Chain Insurance Innovation
Competition is only a testing ground for commercial deployment. The larger opportunity lies beyond the arena. A humanoid robot industry development report released by the China Academy of Information and Communications Technology projected that the whole-machine market for humanoid robots will reach CNY 50 billion by 2035. By 2045, the report projected, the whole-machine market could reach CNY 10 trillion, with more than 100 million robots in use. Those projections point to a vast space for embodied intelligence and for the risk-management systems that must accompany it.
Unlocking that potential requires insurance support across the full chain and the full life cycle. Several leading insurers are accelerating their layout for embodied intelligence and other emerging technologies. They are developing industry-specific insurance products and providing professional underwriting services for robot companies. Their focus is the entire robotics industrial chain, from invention to deployment and from production to after-sales use.
PICC has built a six-system protection framework. The insurance plan covers six dimensions: robot products, technology research and development, supply chain, production operations, liability, and research and development personnel. It offers six types of systematic protection: robot product protection, technology research and development protection, supply chain protection, production operations protection, liability protection, and research and development personnel protection. The framework is intended to address industry risk pain points with professional risk-management capabilities and support the high-quality development of China’s embodied intelligence industry.
In research and development, PICC introduced concept validation and small-trial comprehensive expense loss insurance, as well as pilot validation comprehensive expense loss insurance. These products provide dedicated risk protection for key research and development projects undertaken by technology companies and research institutions. They are designed to help scientific and technological achievements move from a laboratory solution to a production-line product. That transition is often difficult because promising concepts can fail during validation, scale-up or pilot testing. Insurance can reduce the financial shock of such failures and encourage more ambitious experimentation.
In commercial scenarios, PICC’s comprehensive embodied intelligence protection plan covers property damage caused by traditional disaster causes, human operational errors, electrical causes and network causes. It also covers third-party liability caused by accidents. This combination is important because a humanoid robot can damage itself, damage property or injure a person. It can also suffer a failure linked to software, connectivity or human operation. A single incident may involve multiple causes and multiple parties, making integrated coverage more useful than fragmented policies.
4. Commercial Scenarios: Rental Platforms and Repair Cost Relief
One practical example illustrates how insurance can support embodied intelligence commercialization. PICC provides coverage for the first batch of several hundred humanoid robots on a robot rental platform in the Yangtze River Delta region. If a robot is damaged because of a collision or a transmission device fracture, the platform can receive repair cost compensation of up to nearly CNY 400,000 per robot. That coverage addresses the problem of high maintenance costs.
High repair costs are a significant barrier for rental platforms and other asset-heavy users of embodied intelligence. A humanoid robot is a complex system with actuators, sensors, control units, communication modules and structural components. Repair may require specialized parts, skilled technicians and significant downtime. For a rental business, downtime is not only a maintenance expense. It is also lost revenue and reduced customer confidence. Insurance can smooth those costs and make the economics of rental more predictable.
Rental platforms are an important commercial channel for embodied intelligence because they lower the barrier to adoption. Customers can test humanoid robots without making a large capital purchase. They can deploy robots for specific tasks, events or periods. However, rental also spreads risk across many locations and users. A robot may operate in a warehouse one week and a public exhibition the next. It may be handled by different operators with different levels of training. Insurance helps manage that distributed risk and supports the rental model as it scales.
5. Flexible Products Across Production, Sales, Rental and Use
Taibao P&C has designed a dedicated insurance product for the commercial application of humanoid robots. The product is referred to as Jizhibao. It connects risk protection across the entire chain of production, sales, rental and use. Using a risk anthropomorphic concept, it provides integrated protection for property damage, third-party personal injury and third-party property damage. Its term is flexible, breaking the traditional limitation of annual policies. It supports daily, weekly and monthly insurance periods and can precisely adapt to diverse commercial scenarios.
That flexibility matters for embodied intelligence because deployment patterns are not uniform. Some robots are used continuously in production. Some are rented for short projects. Some are demonstrated at events. Some are tested in pilot programs. A traditional annual policy may not fit every use case. Daily, weekly and monthly options allow operators to match coverage to actual exposure. This can reduce costs for short-term deployments and make insurance more accessible for small and medium-sized enterprises that want to experiment with embodied intelligence.
Full-chain coverage also reflects the way value moves through the robotics industry. A robot is manufactured, sold or leased, operated and serviced. Risk can arise at each stage. A production defect may appear after delivery. A lessee may damage the robot. A third party may be injured during operation. A software update may introduce new behavior. Insurance that connects these stages can help align incentives and provide continuity. It can also give end users more confidence that problems will be handled without crippling financial loss.
6. New Risk Characteristics Demand New Underwriting Capabilities
Humanoid robots are new, and their risk exposure is unique. Historical claims data is almost blank, which creates challenges for insurers. Traditional underwriting relies on past loss experience, actuarial tables and stable risk categories. Embodied intelligence does not yet offer a deep historical record. It also changes quickly. New models, new software, new sensors and new use cases appear continuously. Insurers must therefore develop new capabilities rather than simply extend old products.
A PICC P&C executive described several new risk characteristics. First, risk sources are composite. A loss may not come only from natural disasters, fire, explosion or hardware failure. It may also come from algorithm bias or system vulnerabilities. Second, responsibility attribution is complex. If an accident occurs, should responsibility be assigned to the hardware manufacturer, the software developer or the user? That determination is far more complex than with traditional mechanical equipment. Third, emerging risks such as cybersecurity, algorithm safety and information leakage are not covered by traditional enterprise property insurance.
- Risk sources are composite, combining physical, electrical, software, algorithmic and network causes.
- Responsibility may be shared among hardware makers, software developers, operators and users.
- Cyber risk, algorithm safety and information leakage require new coverage beyond traditional property insurance.
- Historical claims data is almost absent, making pricing and risk selection more difficult.
- Rapid technology iteration can change the risk profile during the life of a policy.
These characteristics mean that insurance for embodied intelligence cannot be static. It must be dynamic, data-informed and closely connected to engineering reality. Insurers need to understand how robots perceive, decide, move and interact. They need to evaluate not only the machine but also the environment, the operator, the software version and the network connection. That requires collaboration with robotics companies, research institutions and technology experts.
7. Dynamic Assessment and Cross-Sector Collaboration
PICC has adopted a model of dynamic assessment and cross-sector cooperation. At the early stage of the humanoid robot business, the insurer refers to underwriting data accumulated from mature categories such as industrial robots, service robots and special robots. It combines that data with dimensions including humanoid robot safety level, operating scenario and maintenance cost to dynamically improve pricing accuracy. This approach recognizes that humanoid robots are new but not entirely disconnected from earlier robotics experience.
The insurer also goes into laboratories and production workshops. It works deeply with robot manufacturers and joins upstream and downstream enterprises and research institutions to conduct a kind of health check for robots. The risk is decomposed into the robot’s brain, cerebellum and body. This metaphor reflects the layered nature of embodied intelligence. The brain relates to high-level decision-making and artificial intelligence. The cerebellum relates to motion control, balance and coordination. The body relates to actuators, structures, sensors and power systems. By examining each layer, insurers can build a more complete view of potential failure modes.
This collaborative model can improve more than insurance pricing. It can generate feedback for product design, safety standards and operational procedures. When insurers identify recurring risk factors, manufacturers can address them in future models. When operators understand coverage conditions, they can adopt better training and maintenance practices. When claims data accumulates, it can support more accurate underwriting and more targeted risk prevention. In this way, insurance becomes part of a learning system for embodied intelligence.
Cross-sector cooperation is especially important because no single company controls the entire risk. A humanoid robot may include components from multiple suppliers, software from several developers, connectivity from telecom providers, and services from integrators. A failure may emerge from the interaction of these parts rather than from any one component. Insurers can serve as a neutral party that helps map risk across the value chain and allocate responsibility more clearly.
8. Agile Iteration and Modular Design
Ping An P&C has built a model of dynamic adaptation and agile iteration to address emerging risk-management challenges in humanoid robots and embodied intelligence. The company established a new emerging technology risk laboratory composed of technical experts, insurance actuaries and industry researchers. The laboratory continuously tracks robot technology evolution and application trends. This structure allows the insurer to combine technical understanding with actuarial discipline and industry insight.
In product design, Ping An P&C relies on modular and parameterized design. It breaks coverage responsibilities into basic modules and optional modules. This allows the insurer to adjust coverage scope and pricing parameters quickly as technology iterates and application scenarios change. For example, a robot used in a controlled factory may need a different risk profile from a robot used in a public space. A robot performing repetitive tasks may differ from one interacting with patients or elderly users. Modular design helps match coverage to those differences.
The company also sets a technology iteration risk-sharing clause. In the contract, it reserves a mechanism for periodic review and pricing optimization based on actual loss data. That mechanism is intended to ensure the product covers current risks while adapting to future development. This is important because a policy written today may still be in force when the robot receives a major software update or begins a new task. Without flexibility, coverage could become mismatched with reality. With periodic review and risk sharing, insurer and insured can adjust as knowledge grows.
This agile approach reflects a broader shift in insurance for emerging technology. Instead of waiting for years of stable data, insurers are creating products that can evolve. They are combining expert judgment, dynamic assessment, modular structure and contractual flexibility. That shift is particularly relevant for embodied intelligence, where technical change is rapid and use cases are diverse. The goal is not to eliminate uncertainty but to manage it in a way that supports innovation.
9. Why Insurance Is Becoming Infrastructure for Embodied Intelligence
Insurance is often viewed as a financial backstop after an accident. In the case of embodied intelligence, it is also becoming infrastructure for commercialization. When a company considers deploying humanoid robots, it must evaluate not only purchase or rental cost but also liability, downtime, repair and reputational risk. Insurance can convert some of those uncertain costs into predictable premiums. That predictability helps budgets, investment decisions and customer contracts.
For manufacturers, insurance can support product adoption by reducing customer concerns. For rental platforms, it can protect valuable assets and stabilize cash flow. For end users, it can provide protection if a robot causes injury or property damage. For research institutions, it can reduce the financial consequences of failed validation. For the entire embodied intelligence ecosystem, it can encourage more experimentation by limiting downside risk.
Insurance also creates data and feedback loops. Claims reveal what fails, where it fails and why. Underwriting questions encourage companies to document safety procedures and responsibilities. Risk inspections can identify hazards before they cause loss. Over time, this information can inform standards, training, maintenance schedules and product design. The result is not only better insurance but also safer and more reliable embodied intelligence.
This is why the insurance response to humanoid robots is significant. It signals that embodied intelligence is being treated as a real industrial category with real risks and real commercial needs. It also signals that the market is preparing for scale. Pilot projects can proceed with limited protection, but mass deployment requires systems for liability, asset protection and risk transfer. Insurance is one of those systems.
10. Key Coverage Areas in Humanoid Robot and Embodied Intelligence Insurance
| Coverage Area | Example Risks | Policy Forms Described |
|---|---|---|
| Property and physical loss | Hardware destruction, electrical failure, collision, transmission device fracture | Embodied intelligence robot property damage insurance; comprehensive protection plans |
| Third-party liability | Accidental injury to people or damage to third-party property | Embodied intelligence robot third-party liability insurance |
| Artificial intelligence service liability | AI service failures and related service risks | Artificial intelligence service liability insurance |
| Cyber and network risk | Cybersecurity incidents, system vulnerabilities, information leakage | Multi-dimensional event protection; emerging risk coverage |
| Research and pilot testing | Failed concept validation, pilot validation expenses | Concept validation and small-trial comprehensive expense loss insurance; pilot validation comprehensive expense loss insurance |
| Supply chain and production | Supply chain disruption, production operation risks | Six-system protection covering supply chain and production operations |
| Research and development personnel | Risks to research and development personnel | Research and development personnel protection |
| Commercial use | Rental, operation, sales and use-stage risks | Flexible daily, weekly and monthly coverage; full-chain protection across production, sales, rental and use |
The table shows that insurance for embodied intelligence is not a single product. It is a portfolio of protections that can be combined according to the robot’s life stage and use case. A research laboratory may need expense loss protection for validation. A manufacturer may need product and supply chain protection. A rental platform may need property damage and third-party liability. An end user may need flexible coverage that matches short-term deployment. This modular nature makes insurance adaptable to the diverse paths through which embodied intelligence reaches the market.
11. Market Projections and the Scale Challenge
| Projection Year | Humanoid Robot Whole-Machine Market | Robot Count |
|---|---|---|
| 2035 | CNY 50 billion | Not specified in the projection |
| 2045 | CNY 10 trillion | More than 100 million units |
The projections point to a scale challenge. If the market reaches CNY 50 billion by 2035 and CNY 10 trillion by 2045, with more than 100 million robots in use, the number and variety of risk events will grow. Insurers will need to handle many more policies, claims and risk assessments. They will also need to understand risks across many industries and geographies. The products being developed today are therefore not just pilot instruments. They are early building blocks for a future risk-management market around embodied intelligence.
Scale also changes the nature of risk. In small pilots, human oversight may be high and operating conditions tightly controlled. In large deployments, robots may operate in less controlled environments, with less direct supervision, across longer hours and more varied tasks. The probability of rare events may increase simply because there are more robots and more operating hours. Insurance systems must be able to detect patterns, price risk and respond to claims efficiently. That requires data, technology and collaboration.
The insurance industry’s current work with humanoid robot events, rental platforms and manufacturers is therefore a form of early infrastructure building. It is testing products, gathering experience and creating relationships. The lessons learned now will influence how the industry supports embodied intelligence when deployment becomes widespread. The market projections make that preparation necessary rather than optional.
12. The Road Ahead for Embodied Intelligence and Insurance
The road ahead will require continuous innovation. Insurers must keep pace with changes in hardware, software, artificial intelligence, connectivity and human-robot interaction. They must improve their ability to assess algorithm safety, cybersecurity and responsibility allocation. They must work with manufacturers to understand failure modes and with users to understand operating conditions. They must also develop claims processes that can handle complex incidents involving multiple parties and multiple causes.
At the same time, robotics companies can benefit from engaging early with insurers. Clear documentation, safety testing, incident reporting and software version control can improve insurability. They can also help insurers design more accurate products. Collaboration between technologists and underwriters can produce better risk models, better coverage terms and better incentives for safe deployment. This is especially important for embodied intelligence, where the boundary between hardware, software and human action is often blurred.
Flexible products, modular coverage, dynamic pricing and risk-sharing clauses are likely to become more common. They allow insurance to evolve alongside technology rather than lag behind it. They also allow different users to select protection that matches their exposure. A startup testing a prototype may need different coverage from a logistics company operating a fleet. A hospital may need different protection from a factory. A rental platform may need different terms from an end user. The ability to customize coverage is central to serving the embodied intelligence market.
The broader significance is that insurance can help bridge the gap between technical possibility and commercial reality. A robot may be technically capable but commercially difficult if its risks cannot be managed. Insurance helps make risk visible, transferable and priceable. It supports trust between manufacturers, operators and customers. It can also encourage responsible development by tying coverage to safety practices and data transparency. In this sense, insurance is not separate from innovation. It is part of the ecosystem that allows innovation to scale.
13. What the Insurance Shift Signals for Embodied Intelligence
The movement of insurers into humanoid robot and embodied intelligence coverage signals several things. First, embodied intelligence is no longer seen only as a research topic. It is being treated as an emerging industry with assets, liabilities, operations and commercial relationships. Second, the risks are being taken seriously. Companies are not assuming that existing machinery insurance can simply be renamed. They are designing new products for new risk characteristics. Third, collaboration is becoming essential. Insurers, robot makers, software developers, research institutions and users must work together to understand risk.
Fourth, insurance is becoming a factor in business model design. Rental platforms, robotics-as-a-service providers and technology integrators need coverage that matches their cash flow and asset exposure. Flexible daily, weekly and monthly policies support new business models. Full-chain coverage supports companies that handle multiple stages of the robot life cycle. Expense loss insurance supports research and validation. These products can influence which business models are viable and how quickly they scale.
Fifth, the development of insurance for embodied intelligence can contribute to broader safety and standards discussions. Claims data and risk assessments can reveal where accidents happen and what causes them. That information can inform design improvements, operator training and safety guidelines. While insurance cannot eliminate risk, it can make risk more transparent. Transparency is a foundation for trust, and trust is essential if humanoid robots are to work alongside people in factories, warehouses, hospitals and homes.
The Second World Humanoid Robot Games showed how far embodied intelligence has come in movement and control. The insurance activity around the event and beyond shows how the supporting ecosystem is developing. Both are necessary for the transition from laboratory demonstrations to sustained commercial deployment. The race is not only about building robots that can move. It is also about building systems that allow those robots to operate safely, reliably and at scale.
14. Conclusion: Insurance as a Bridge from Laboratory to Deployment
Humanoid robots are moving from competition arenas into factories, warehouses, healthcare settings and elderly care. Embodied intelligence is moving from laboratory prototypes toward commercial services. That journey depends on technical progress, but it also depends on risk management. Insurance companies are developing products that cover property damage, third-party liability, artificial intelligence service liability, cybersecurity, research and pilot testing, supply chain, production operations, liability and research personnel. These products are being adapted to the unique characteristics of humanoid robots and embodied intelligence.
PICC has built event protection for the Second World Humanoid Robot Games and a six-system framework for the broader robotics industry. Taibao P&C has designed a flexible product across production, sales, rental and use. Ping An P&C has developed a dynamic adaptation and agile iteration model with modular design and risk-sharing clauses. Together, these efforts show that insurance is preparing for a future in which embodied intelligence is widely deployed. The market projections of CNY 50 billion by 2035 and CNY 10 trillion by 2045, with more than 100 million robots in use, highlight why that preparation matters.
The path ahead will not be simple. Historical claims data is limited, risk sources are composite, responsibility is complex, and technology changes quickly. Yet these challenges are also opportunities for innovation. By combining dynamic assessment, cross-sector collaboration, modular products and flexible terms, insurers can help companies manage the financial consequences of risk. In doing so, they can support the safe and sustainable growth of embodied intelligence. Insurance will not make every robot perfect, but it can help make the transition from laboratory to real-world deployment more orderly, more trustworthy and more commercially viable.