At the World Humanoid Robot Games in China last month, humanoid machines jumped, boxed and danced before cheering spectators. Some fleet-footed robots ran faster than Usain Bolt over 100 metres. Their increasingly capable performances attracted attention not only for what the machines could do, but also for their spectacular and sometimes meme-worthy stumbles.
For China’s military establishment, however, the significance of the event extended well beyond entertainment.
Two days after the Games ended, the People’s Liberation Army’s official newspaper, PLA Daily, drew its own conclusions. It called for researchers to accelerate the transfer of cutting-edge technologies from laboratories to military training grounds for robotic “combatants”.
The shift reflects a broader effort under way inside China’s defence establishment. A Reuters review of more than 100 Chinese military procurement notices, academic studies, patents, official publications, government records and defence-company materials found that Chinese military institutions are examining how humanoid robots could eventually be used in war.
The research gained momentum in 2025 and 2026. Military institutions have been testing humanoids against battlefield requirements, seeking robots and technologies to train them, and examining how the machines could operate alongside troops. Much of the work has centred on perception, manipulation and training data — the systems that could determine whether humanoid robots can move beyond controlled demonstrations and function in unpredictable environments.
China’s commercial strength gives its military an unusual industrial base from which to pursue those ambitions. Chinese manufacturers accounted for about 95% of global humanoid shipments in 2025, according to BofA Global Research.
Yet the gap between technological demonstrations and actual warfare remains substantial. Reuters found no evidence that China has deployed an armed humanoid with an operational PLA unit. The machines remain energy-intensive and unreliable outside controlled settings. Battlefield environments introduce variables that can challenge robots’ abilities in ways that demonstrations cannot easily replicate.
Even so, Chinese military research is already imagining how humanoids might be integrated into combat.
One scenario described in an August 2025 paper by researchers at the National University of Defense Technology’s test centre in Xi’an envisaged an urban assault force consisting of six “combat robots” — humanoids, robot dogs or unmanned vehicles — divided between two assault teams.
Working alongside ground troops, the robots would clear an enemy-occupied building floor by floor and room by room. The researchers modelled the scenario using equipment they projected could become available within five to 10 years. The study did not specify the robots’ rules of engagement, the weapons they would carry or how they would respond to encounters with civilians.
The attraction of humanoids lies partly in their physical resemblance to the environments built for people. Improvements in balance, perception and endurance could make them useful in buildings, tunnels and ship interiors. Their two hands and two feet could, in theory, combine mobility, climbing and manipulation.
But humanoids are not necessarily the most practical machines for every military task. For reconnaissance and logistics, quadruped, tracked or wheeled robots would generally be cheaper and easier to deploy at scale.
Chinese military commentary nevertheless points towards increasingly ambitious roles. A July 2025 PLA Daily article argued that humanoid missions could eventually develop “from supporting combat toward primary combat”. It also discussed circumstances in which robots might be authorised to fire on a living target that had first been verified by a human.
In December, the PLA Eastern Theater Command released an AI-generated montage depicting military robots, including humanoids and quadrupeds, overwhelming Taiwan’s defences in a hypothetical future conflict.
The exploration of military humanoids is not confined to China. Last year, the U.S. Army launched a competition aimed at developing “militarized humanoid capabilities” that could eventually operate alongside soldiers. Proposed roles included reconnaissance, security, obstacle clearing, hazardous-material operations and offensive and defensive missions in urban terrain.
Yet China’s advantage in commercial robotics remains significant. The country’s manufacturers dominate humanoid shipments, while its defence institutions are increasingly examining how commercially developed technologies might be adapted for military purposes.
The Chinese military’s interest became more pronounced in 2024, when the National University of Defense Technology and China’s Academy of Military Sciences hosted a defence-technology forum that included a session on military applications for humanoid systems.
By the following year, the concept had moved closer to battlefield experimentation.
A university competition organised in June 2025 included an event simulating “infiltration behind enemy lines”. Robots were required to enter an area, map it and locate targets. Other scenarios imagined robots conducting identity checks, discipline inspections and security patrols at military bases.
The military-research university described the competition partly as a proving ground for humanoids to eventually “move toward the battlefield”.
Procurement activity around the same period suggested that the ambition was not merely theoretical. In May 2025, the PLA issued a tender for a humanoid robot, including installation and technical training. Four months later, a procurement notice listing an NUDT email as its contact address sought an “embodied humanoid robot intelligent perception and dexterous operation system”.
The available records do not establish whether the tenders were fulfilled, which companies might have won them or which robot models may have been supplied.
In June this year, another military procurement document showed the PLA budgeting about $300,000 for a system to collect and label camera, radar and motion data for humanoids. The data would include information about terrain that robots could traverse.
The focus on perception, manipulation and training suggests that China’s military is looking beyond the physical robot itself. The challenge is not simply building a machine that can walk, but creating systems capable of understanding surroundings, manipulating objects and responding effectively to complex environments.
By 2026, the development had also spread into China’s defence industry.
Norinco, the state-owned defence conglomerate, has developed the Fuxi robot, a full-size humanoid that company materials published in August described as capable of sentry duty, all-weather reconnaissance, intelligent patrol and dangerous roles.
Fuxi can also be paired with Norinco’s teleoperation system, allowing a human operator to control the humanoid remotely. Such technology could reduce the requirement for a robot to make every decision autonomously.
But remote operation does not eliminate the deeper questions surrounding the eventual use of autonomous machines in war.
Military systems capable of operating independently would have to confront situations involving civilians, wounded combatants and fighters attempting to surrender. The laws of war require forces to distinguish combatants from civilians and refrain from attacking wounded or surrendering fighters. The International Committee of the Red Cross has warned that autonomous weapons could struggle to interpret signs of surrender because such decisions may depend on context, conduct and intent.
The U.S. military requires autonomous weapons to undergo legal review and realistic testing, while commanders and operators must exercise appropriate judgement over the use of force. China’s defence ministry said in March that AI-powered weapons should remain under human control.
For now, the limitations of humanoid robots remain considerable. Malcolm Davis, a senior analyst at the Australian Strategic Policy Institute, said he did not regard them as practical battlefield systems at present. But he added that they could become so “in five to 10 years”.
China’s own researchers acknowledge the obstacles. Wang Yonghua, a researcher at the Academy of Military Sciences, wrote in a November commentary that humanoids still face unresolved problems involving movement, perception and intelligence. Technical breakthroughs, lower manufacturing costs and industrial-scale production, he argued, would be necessary before humanoids could become widespread military systems.
If those conditions are eventually met, Wang wrote, “humanoid robots will stream onto the battlefield.”
The journey from robots dancing before cheering crowds to machines operating alongside soldiers is therefore still unfinished. But China’s military research shows that the distance between the two is now being treated not simply as a technological curiosity, but as a problem of military planning.
