$1.5 Billion and 10 Years Produced 30 Brain Implant Patients. China Just Approved a Commercial Device Anyway.
On March 13, 2026, China’s pharmaceutical regulator approved the world’s first commercial brain-computer interface. Combined, Western BCI companies have spent $1.4 billion to implant 22 patients in clinical trials. A first-ever global patient census reveals the most expensive per-person technology program since the space race.
Twenty-two. That is the total number of people in the United States and United Kingdom who have a functioning brain-computer interface implanted in their skulls right now. Combined, the companies that put those devices there have raised approximately $1.4 billion in venture capital.
That works out to roughly $64 million per patient.
On March 13, 2026, while those 22 patients continued participating in clinical trials with no guaranteed commercial endpoint, China’s National Medical Products Administration (NMPA) quietly approved the world’s first commercial brain-computer interface for clinical use. The device, built by Borui Kang Medical Technology (operating as Neuracle) of Shanghai, is now available for purchase by hospitals. Not in a trial. Not under an investigational exemption. As a product.
And the reason this milestone received so little initial Western coverage illuminates something uncomfortable about the gap between American innovation and Chinese deployment.
What Borui Kang Built
Borui Kang’s system is an implantable brain-computer interface designed for a single, specific function: enabling quadriplegic patients to grasp objects using a pneumatic glove. Its implant sits in the epidural space, on the outside of the brain’s protective membrane. It reads electrical signals through the dura mater, transmits them wirelessly to external hardware, and a decoding algorithm translates the signals into commands for a glove that physically opens and closes the patient’s hand.
Seven components make up the system: an implantable EEG electrode kit, a BCI implant, an EEG transceiver, a pneumatic glove, a disposable surgical tool kit, EEG decoding software, and clinical management software. It is indicated for adults aged 18 to 60 with quadriplegia caused by C2-C6 cervical spinal cord injury graded A through C on the ASIA Impairment Scale.
Placed through a minimally invasive procedure, the device avoids open brain surgery entirely. This is the critical design choice that made commercial approval possible.
The Global Patient Census
To understand what China just did, you need to see the full picture of where the BCI industry actually stands. I compiled patient counts across every major implantable BCI program worldwide, using company announcements, regulatory filings, and hospital disclosures. No one, to my knowledge, has published this composite before.
| Company | Country | Approach | Patients | Total Funding | Status |
|---|---|---|---|---|---|
| Neuralink | US/UK | Intracortical (1,024 electrodes) | 12 | ~$1.1B | Early feasibility study |
| Synchron | US/AU | Endovascular (Stentrode) | ~10 | ~$275M | IDE clinical trial |
| Precision Neuroscience | US | Cortical surface (Layer 7) | ~3 | ~$100M | Acute recordings |
| CorTec | Germany | Cortical (AirRay) | ~3 | ~$40M | Compassionate use |
| Borui Kang (Neuracle) | China | Epidural (EEG) | Undisclosed | Undisclosed | Commercially approved |
Global total: approximately 28 to 30 people worldwide have a functioning implantable BCI. Combined investment to reach this point exceeds $1.5 billion. That is roughly $50 million per patient, making BCI the most expensive per-capita medical technology program in history. For comparison, the Human Genome Project cost $2.7 billion and decoded 3 billion base pairs relevant to every human on Earth. By comparison, BCI has spent half that amount and served fewer people than fill a single school bus.
Neuralink’s numbers tell the starkest story. Twelve patients, $1.1 billion in funding, and a $9.6 billion valuation. That is $92 million per implanted patient, or $800 million per billion dollars of valuation for every person who can actually use the product. Across those 12 patients, more than 15,000 hours and 2,000 days of device use have been logged. Seven of them received implants at University College London Hospitals between October and December 2025 as part of the GB-PRIME study, with one patient, medical student Sebastian Gomez, reporting he could control a computer and phone through thought alone within hours of surgery.
Synchron’s approach is cheaper per patient ($27.5 million each, approximately) and avoids open brain surgery by threading its Stentrode device through blood vessels to the motor cortex. About 10 patients in the US and Australia have received implants. In May 2025, Synchron became the first BCI company to achieve native Apple integration, letting patients control iPhones, iPads, and Apple Vision Pro with thought alone. Synchron raised $200 million in a Series D in November 2025, with backing from Gates and Bezos.
Two Regulatory Philosophies
The divergence between China and the United States on BCI regulation is not about speed. It is about scope.
FDA rules require BCI companies to move through a sequence of gates: Investigational Device Exemption (IDE) for early feasibility studies, then confirmatory trials, then a Pre-Market Approval (PMA) application or De Novo classification. Neuralink received its IDE in May 2023 and implanted its first patient in January 2024. At the current pace, with 12 patients over 27 months, a full trial of 50 to 100 patients could take until 2029 or 2030, followed by a PMA review of 12 to 18 months. Best-case commercial availability: 2031.
China’s NMPA approved Borui Kang’s device on what appears to be a faster but narrower track. The approval covers one specific indication (cervical spinal cord injury, grades A-C), one specific function (grasping via pneumatic glove), and one specific placement (epidural). Borui Kang did not try to build a general-purpose brain interface. They built a grasping tool and got it through the regulatory door.
This is not recklessness. It is the same philosophical split that separated Chinese and American approaches to high-speed rail, 5G deployment, and EV adoption. China optimizes for deployment at scale on proven-enough technology. The US optimizes for exhaustive validation of ambitious technology. Both approaches have failure modes. China’s risk is deploying a device with limited post-market safety data. America’s risk is that patients who could benefit today wait a decade for a device that is 10 times better.
The Parking Lot Problem
Here is the strongest case against treating China’s approval as a genuine lead.
An epidural BCI reading signals through the dura mater is like trying to listen to a symphony from the parking lot. You can hear the bass drum and maybe pick out whether the orchestra is playing loud or soft, but you will never distinguish the first violin from the second. Borui Kang’s device captures enough signal to differentiate a grasping intention from rest. Neuralink’s 1,024 intracortical electrodes, threaded directly into the brain tissue by a surgical robot, can decode individual finger movements, mouse cursor trajectories, video game controls, 3D design inputs, and, in at least one ALS patient, conversational speech.
In capability terms, the gap is enormous. Borui Kang’s patients can squeeze a glove. Neuralink’s patients can browse the web, text friends, and design machine parts with thought alone. These are not the same product class, and calling them both “brain-computer interfaces” is about as useful as calling both a bicycle and a fighter jet “vehicles.”
The counterargument to the counterargument: Borui Kang’s patients can do something. Right now. Outside a clinical trial. That matters to the roughly 250,000 to 500,000 people who sustain spinal cord injuries globally every year, most of whom will never qualify for a Neuralink trial.
China Is Not Stopping Here
The Borui Kang approval exists within a larger Chinese BCI strategy. Beijing’s 15th Five-Year Plan explicitly names brain-computer interfaces as a key emerging industry. NeuCyber NeuroTech, a Beijing-backed startup, openly told Reuters in March 2026 that it is “three years behind Neuralink” on intracortical technology, which is notable for two reasons: they’re being honest, and they’re announcing a timeline.
Some Chinese provinces have already begun piloting BCI treatments under insurance programs, according to Reuters reporting from March 2026. If reimbursement infrastructure scales faster than device capability, China could create a market that pulls technology forward through demand rather than pushing it through research grants. That is how Chinese solar panel manufacturing went from a niche industry to 80% of global capacity in 15 years.
Limitations
This analysis has meaningful gaps. Borui Kang has not disclosed its clinical trial patient count, adverse event data, or detailed functional outcomes publicly. Borui Kang’s device pricing and hospital rollout plans are unknown. My cost-per-patient calculations divide total funding by total implanted patients, which conflates R&D investment (the vast majority of spending) with per-procedure costs. The actual surgical cost of a Neuralink implant is estimated at $40,000 to $250,000 depending on facility and follow-up care, not $92 million. Synchron and Precision Neuroscience patient counts are approximate, based on press reports rather than regulatory filings. And because Borui Kang’s device is epidural rather than intracortical, the useful comparison set is more limited than the table above implies.
What You Can Do
If you are a neurologist or rehabilitation physician: Track the NMPA approval closely. The Borui Kang device represents the first commercially available option for spinal cord injury patients outside a trial. If your institution treats SCI patients and has connections to Chinese medical device distributors, inquire about procurement pathways. The regulatory precedent also accelerates timelines for epidural BCI devices in other jurisdictions. The European Medicines Agency and FDA both accept comparative regulatory data.
If you are an investor: The per-patient math should give you pause. At $92 million per patient, Neuralink needs 100,000 patients at $50,000 each just to recoup its current funding. That is a $5 billion revenue target before a single device has been priced, insured, or sold. Synchron’s endovascular approach has a clearer path to scale because it uses existing catheterization labs rather than requiring a custom surgical robot. Watch their confirmatory trial enrollment numbers in 2026.
If you or someone you know has a spinal cord injury: Options are expanding faster than most clinicians realize. The ClinicalTrials.gov registry lists active BCI studies by indication and location. Neuralink’s GB-PRIME study at UCLH is enrolling in the UK. Synchron’s COMMAND trial is active in the US. Neither requires you to be near a major research hospital for initial screening. For the Borui Kang device in China, contact major Chinese neurosurgery centers directly, as commercial rollout details have not been published in English.
The Bottom Line
The brain-computer interface industry just split into two races. In the first race, companies like Neuralink and Synchron are building devices that could eventually let paralyzed people do everything a non-disabled person does with a computer. That race is somewhere in its third lap of a ten-lap course, with 22 patients and $1.4 billion spent. In the second race, a Chinese company built a device that does one thing and got it into hospitals. The patients in that second race cannot browse the web or play video games with their thoughts. They can pick up a cup. For someone who has not held anything in years, the cup is enough.