Robots Are Coming — but Not Everywhere

Getty Images “The ChatGPT moment for robotics is coming,” declared Nvidia CEO Jensen Huang at the Consumer Electronics Show in January 2025. It’s a widespread expectation: that humanoid robots will follow the same explosive adoption curve as generative AI. Our research suggests the opposite. Humanoid robotics will be adopted unevenly, across diverging use cases and geographies. Leaders need to reset their strategies for a more complex, jagged path to scale. Humanoid robots are developing at pace. Engineering advances and new powerful models of physical intelligence, including vision-language-action models and improved locomotion systems, are converging to create a robotics super-cycle. But leaders who expect the path of humanoid adoption to mirror that of generative AI are misjudging the strategy required. A Less Predictable Pace Our research shows that three fundamental forces are shaping a very different adoption curve. Force 1: No One-Size-Fits-All Model Some humanoid manufacturers envision general-purpose robots working across multiple activities and industries. But for the foreseeable future, humanoid robots will not be commercially deployed as universal productivity tools. Consider the multitude of roles that humanoids could theoretically take on: carer, guard, soldier, inventory picker, assembly line worker, hospital assistant, or concierge, to name but a few. Each role, and often each separate task of that role, imposes distinct and often incompatible requirements. A warehouse robot, for example, may need to lift up to 132 pounds, requiring sophisticated actuators — the equivalent of human muscles — to complete its task successfully. Those actuator systems will make up 40% to 60% of its cost. But a care assistant robot requires different capabilities: subtle facial expression, fine motor control, and emotionally sensitive interaction. In this case, the technology build will tilt more heavily toward perception and haptic technologies. The diversity of humanoid roles implies fundamentally different technology build specifications, operating requirements, and ultimately, business cases for deployment. This divergence extends beyond the physical build into software and connectivity. A security robot will use low-latency edge-computing data processing close to its physical location — so it can rapidly respond to environmental changes, such as an incursion by an unknown person or vehicle. By contrast, a humanoid hospital assistant may rely on large graphical models to navigate hospital facilities and create 3D images of patient charts and X-rays. Each scenario requires different computing architectures. This specialization, in turn, supports an array of humanoid variations, such as fully mobile bipedal robots or wheeled torsos, reflecting the different physical demands of each role. The result is the de facto specialization of competing humanoid manufacturers, each prioritizing different market niches. Analysis of these companies shows heavy focus by market segment — manufacturing, warehousing, care, service, and home — in market strategy and customer cases. Humanoid robots will therefore come in a variety of shapes and specifications. They are being built as specialists, not generalists, and will be scaled role by role. Force 2: Geographic “Islands of Viability” Robot costs are dropping steadily, but every humanoid deployment must be justified with a clear return on investment — a determination that will vary sharply by location due to workplace regulations, labor market dynamics, and demographics, among other factors. Consequently, early humanoid business cases will be concentrated in a few “islands of viability” where conditions are ripest for first deployments. ROI will be strongest in locations that have high labor costs and acute workforce shortages, or that have existing infrastructure and potential for immediate mass-market scale. There will be fewer compelling business cases in parts of the world with abundant and cheap labor, or limited humanoid manufacturing ecosystems. Japan is arguably the world’s most pronounced island of robot viability for care and manufacturing tasks. Facing a projected 11 million worker shortfall by 2040 , the Japanese Ministry of Economy announced new procurement targets to drive robotics adoption and recently announced it would provide nearly 400 billion yen (approximately $2.4 billion) in aid to build an AI system to control robots. South Korea faces similar demographic pressures, with a rapidly aging population and the world’s lowest birth rate. Both countries are seeing rapid humanoid investment. For instance, Hyundai plans to invest $6.3 billion to build a new robotics manufacturing complex in South Korea as the company shifts its strategy from automotive manufacturing. Other islands of viability are centered around manufacturing. Germany has a documented skills shortage, while Singapore has a limited domestic labor pool and strict foreign labor quotas — conditions in which humanoid deployment can generate returns for companies deploying them. China presents a different dynamic. Its full-stack supply chain, competitive manufacturing base, and mass-market scale allow manufacturers to build humanoids at a fraction of the costs faced by competitors. It is estimated that China built 90% of the humanoid robots produced globally in 2025. Two companies, Agibot (based in Shanghai) and Unitree Robotics (in Hangzhou), together shipped over 10,000 units in 2025, while U.S.-based peers like Figure AI and Tesla remained in the low hundreds for actual customer deliveries . The Lunar New Year celebrations in 2026 produced by China Central Television, featuring dancing Unitree robots , was a national declaration of intent for this sector, reflecting long-term government investments. Regulation adds another layer of geographic complexity, as countries take divergent approaches to safety, privacy, and liability, especially in sensitive settings such as health and social care. As with semiconductors and AI software, humanoid robotics will likely be entwined with the geopolitics of technology, which could result in parallel humanoid ecosystems in the U.S. and China. The message is clear: For years to come, where you locate humanoids will be fundamental to their viability. Force 3: The Human Factor The third factor shaping humanoid adoption is the least understood and potentially the most consequential. People are physiologically wired to react to the human form. When a humanoid enters a workspace, care home, or shop floor, its human resemblance triggers instinctive and visceral responses that other technologies don’t. Human reaction to humanoids will be an unpredictable but crucial factor in the success of their rollout. Most attention focuses on the potentially negative user reactions of humanlike robots. The uncanny valley effect, first identified in 1970 by Japanese researcher Masahiro Mori , captures how encounters with humanlike robots can leave people unsettled and uneasy. Yet the full picture is more complex. A growing body of research investigates the diverse and often polarized reactions that humanoids engender. For example, in care settings, a 2025 University of Cambridge study found informal caregivers became increasingly comfortable sharing their emotional struggles with a humanoid robot, and subsequently experienced measurable improvements in loneliness and mood. Some research points to more ambivalent reactions; one study in the hospitality industry found that while two-thirds of guests positively reviewed experiences with robot servers, 28% reported discomfort — with little middle ground. Human personality types also affect the quality and effectiveness of human-robot interactions. Emerging research suggests that humans tend to prefer humanoid robots with similar personality types to their own. Put simply, extroverts will likely prefer humanoids with more social personalities or exaggerated behaviors, while int