China approves first brain implant for paralysed patients – what’s behind the decision

China is now making this a reality.

China’s medicines regulator has, for the first time, approved a brain implant for routine use in patients living with paralysis. The system is designed to convert signals from the brain into usable hand movements - a development that could significantly reshape the race for the future of neurotechnology.

How the new NEO brain implant works

The platform is called NEO and was developed by Shanghai-based medical technology company Neuracle. At its core is a round, wireless chip about the size of a coin, fixed to the outer membrane covering the brain above the motor cortex. To place it, surgeons open the skull and expose the brain’s surface, but they do not penetrate the brain tissue itself.

When the user tries to move their hand - even if a spinal cord injury prevents the muscles from responding - the chip records the electrical activity generated by that intention. On its own, these raw signals would be difficult for outsiders to interpret, so dedicated software converts them into precise control commands.

"Thought in the head, movement in the glove: NEO translates neural activity into mechanical force."

Those commands are then sent to a specialist glove. The glove contains small air chambers that inflate or deflate using compressed air. This creates controlled, artificial finger opening and closing - making it possible to grasp everyday items such as bottles, cups, or a smartphone.

Why “non-penetrating” still involves surgery

Neuracle describes the approach as “non-penetrating”. In practice, this means the electrodes do not pierce deep into the brain; they sit on the surface. That sets NEO apart from systems that push tiny needle-like electrodes directly into the brain tissue.

• Surface contact rather than deep needles
  
• Lower risk of direct brain injury
  
• Still delivers signal quality high enough for accurate control
  

However, it is not a procedure you can do without an operation. Implantation requires neurosurgery under a general anaesthetic, with the standard risks that entails - including bleeding, infection, or wound-healing complications. Once implanted, the device remains in the head long term and communicates wirelessly with the external system.

China moves ahead of the United States

In March 2026, China’s national medical devices authority granted the NEO system the highest approval category. This allows the implant to be sold and used routinely in suitable patients. China is therefore the first country to allow an invasive brain–computer interface to move from research studies into standard clinical care.

In the United States, start-ups such as Elon Musk’s Neuralink have been developing comparable systems for years. Neuralink has already tested implants in human volunteers - specialist media reports suggest participation has so far reached a little over two dozen people. Even so, there is not yet an official market authorisation in the US.

"With the commercial clearance of NEO, China is sending a signal: neurotechnology should move quickly from the lab to patients’ bedsides."

Momentum is building inside China as well. Companies such as Shanghai NeuroXess have showcased, in pilot studies, paralysed men who were able to control computers and other devices using thought alone just days after surgery. Beijing has also placed brain–computer interfaces into strategic plans as a future-facing field and is promising accelerated approval pathways.

What is driving the global race

Brain–computer interfaces - often abbreviated to BCI - are widely seen as a key technology bridging medicine and high-tech engineering. They are intended to give people living with paralysis, stroke, or neurological conditions new functional abilities. At the same time, some companies are pursuing non-medical uses too, such as “thought-controlled” computers or prosthetics.

Many of today’s approaches trace their roots back to US research programmes such as BrainGate in the 2000s. Teams demonstrated that paralysed individuals could steer a cursor or robotic gripper simply by imagining movements. Companies such as Neuracle are now building on this foundation with the aim of delivering everyday products - not just laboratory demonstrations.

Who the implant is for - and who it is not

China’s approval for NEO applies to a clearly defined group. The target population is adults aged 18 to 60 with a spinal cord injury in the cervical (neck) region. The injury must be longstanding - at least one year - and the person’s condition should not have changed significantly over the past six months.

Another requirement is that the individual can still lift or move their arms to some extent, but has little or no purposeful grasp. In these cases, a robot-assisted glove can restore meaningful day-to-day capability: holding a cup, controlling a toothbrush, or opening a drawer.

Criterion	Requirement for the NEO system
Age	18 to 60 years
Injury	Damage to the spinal cord in the cervical region
Duration of paralysis	At least 12 months
Stability	No significant change for 6 months
Residual movement	Some arm movement remains, but grip is severely limited

In clinical testing, participants’ grasping ability improved in measurable ways. They could hold objects more reliably and place them more accurately. The aim is not a miraculous cure, but practical functional gains that make everyday life slightly more independent.

Risks, complications, and open questions

Brain chips may sound futuristic, but they come with a set of medical and technical challenges. The operation itself carries the familiar risk profile of neurosurgery. After implantation, infections can occur, fluid may build up, or the implant could shift position.

Over the longer term, the body often forms scar tissue around foreign objects. This layer can dampen electrical signals and reduce measurement quality. Some systems become less precise over the years as a result. Engineering teams try to offset this using new materials and more sophisticated signal processing.

There are also ethical and legal issues to resolve: Who is responsible if the system outputs incorrect commands? How will highly sensitive brain data be protected? And how long should a chip be allowed - or required - to remain in someone’s head if it stops working or becomes obsolete?

What this development could mean for patients

A market authorisation in China enables something research labs struggle to achieve: real-world data. If hundreds or thousands of paralysed people use NEO in daily life, it will generate large volumes of information on when the system performs well, when it does not, and how it fails. That feedback could make future generations of implants significantly more robust.

For those affected, the focus is on tangible improvements: needing less help getting dressed, drinking independently, and operating a wheelchair, tablet, or light switches. Small increases in autonomy can determine whether someone needs round-the-clock care or can act independently in certain situations.

• Greater independence in everyday life
  
• Reduced burden on relatives and care staff
  
• Possible return to work or education for some people
  

At the same time, questions of cost and access remain. China will now have to decide whether health insurers or the state will pay for these procedures, and how many hospitals can build the necessary expertise. Western health systems will also watch closely to assess whether the benefits justify the risks and expenditure.

How far the journey to Europe and Germany still is

For patients in the German-speaking world, NEO remains a distant prospect for now. Each region has its own approval processes, standards, and data-protection rules. Manufacturers would need to run studies under European requirements and provide years of safety follow-up. Some experts expect comparable options in Europe only over the longer term.

Even so, China’s approval illustrates where medicine could be heading: away from the classic wheelchair alone and towards hybrid systems combining human capability, machines, and software. Anyone living with severe paralysis today may see far more therapeutic options in the coming years - from exoskeletons to thought-based control of assistive devices.

For the moment, neuroimplants such as NEO are likely to remain a niche option for a tightly defined patient group. Yet they mark a turning point: for the first time, a country is selling a brain interface not only for research, but as an officially approved medical product. How safe, useful, and widely accepted this technology truly is will now be tested in everyday life for many of those affected.