China’s Brain-Computer Interface Just Beat Neuralink — Here’s How

The global race for commercial brain-computer interface dominance: China’s NEO approval in March 2026 marked the first time any country commercially cleared an invasive brain-computer interface for sale beyond clinical trials.

*A coin-sized chip. A paralyzed man. And a milestone that quietly changed the future of medicine.*

I have been covering emerging neurotechnology for years. I have read the trials, the regulatory filings, the IPO prospectuses. But when China approved the NEO brain chip on March 13, 2026, something felt different. Not because it was surprising. Because it was inevitable, and almost nobody outside the science community noticed.

Let me tell you what actually happened.

The Moment That Changed Everything

A man named Dong Hui raised a cup to his lips and drank from it. He had not been able to do that without help for over a decade. The device that made it possible was a coin-sized titanium implant sitting on the outer membrane of his brain, reading his motor intention signals and sending them to a pneumatic glove wrapped around his hand.

That device, called NEO, became the world’s first commercially approved invasive brain computer interface on March 13, 2026. Not a trial. Not a prototype. A product a physician can now actually prescribe.

And Elon Musk’s Neuralink? Still waiting for FDA commercial approval. Earliest projection: 2029.

What a Brain Computer Interface Actually Does

Most people imagine brain implants as science fiction. The reality is more precise and more interesting.

AI-generated educational infographic illustrating how a Brain-Computer Interface (BCI) works, from decoding brain signals to controlling external devices. Created to help readers understand the technology through a simplified visual explanation.

A brain-computer interface intercepts the electrical signals your motor cortex generates when you intend to move. For someone with a spinal cord injury, those signals never reach the muscles. The BCI catches them, decodes them using machine learning, and reroutes them to an external device that completes the movement instead.

NEO does this with eight sensors placed on the dura mater, the membrane covering the brain, without penetrating cortical tissue at all. That single design decision is what got it approved first. Lower surgical risk, cleaner regulatory path, faster timeline.

Neuralink uses 1,024 electrode threads inserted directly into brain tissue for far higher signal resolution. The tradeoff is a more invasive procedure and, in the case of their first patient Noland Arbaugh, approximately 85 percent thread retachment within months of implantation.

These are not competing products. They are different bets on the same problem.

Why China Got There First

This is the part that matters beyond the device itself.

China did not just approve a brain computer interface. China built the entire infrastructure around it before the product even existed. The National Healthcare Security Administration established insurance coding categories for invasive BCI procedures a full year before any device was ready. When NEO got approved, insurance coding followed in 12 days.

Beijing named brain interface technology a key industry of the future in its 2026 to 2030 Five-Year Plan, published the same day as the NMPA approval. That is not coincidence. That is industrial policy executed with precision.

In the United States, 374,000 people are living with spinal cord injuries right now. There is no commercially approved invasive BCI available to any of them. The regulatory gap is real, and it has a human cost.

What the Critics Are Right About

I want to be honest here, because the approval is not without legitimate questions.

AI-generated educational comparison illustrating several Brain-Computer Interface (BCI) technologies and neural implant concepts. The image is intended to help readers understand the different approaches being explored in the field of brain-computer interfaces and should be viewed as a conceptual representation rather than actual product photographs.

The primary clinical evidence for NEO is a medRxiv preprint from October 2025 that has not yet completed peer review. The trial covered 36 patients over 18 months. That is meaningful data, but it is substantially smaller and shorter than what the FDA typically requires for a Class III invasive neural device.

Professor Miguel Santos of UCL’s Neural Engineering group put it plainly: scaling BCI to broader populations demands rigorous comparative trials, not just feasibility studies. Brain implant signal degradation from glial scarring remains one of the most persistent unsolved problems across all bci devices currently in development. And five year durability data does not exist for any commercially deployed device anywhere in the world.

Approval is a beginning. Not a verdict.

What This Means for Real Patients

For someone with ALS, locked in syndrome, or a cervical spinal cord injury, this approval is not an abstract milestone. It is the difference between a technology that exists in a laboratory and one that a doctor can prescribe.

NEO is approved only for adults aged 18 to 60 with C2 to C6 cervical injuries who retain some upper arm movement. That is a narrow population. But it is a real one. And every approval like this expands what is possible for the next application, stroke rehabilitation, speech restoration, communication for patients who have lost all voluntary movement.

BrainGate’s decades of research already documented real time voice synthesis from an ALS patient’s implanted electrodes in 2025. The pipeline is moving faster than most people realize.

The Question Nobody Wants to Answer

When the technology works well enough for healthy people, and Elon Musk has said publicly that Neuralink’s long term vision includes exactly that, who owns the neural data a brain computer interface collects?

No regulator anywhere has answered that question. Not the FDA. Not the NMPA. Not the EU. A brain machine interface is simultaneously a medical device and a data collection system embedded in the most sensitive biological structure humans possess. The ethical frameworks have not kept pace with the engineering.

That gap will matter more with every approval that follows NEO.

Where This Goes Next

Chen Liang, a neurosurgeon at Huashan Hospital, has confirmed plans to begin brain computer interface trials on stroke patients before the end of 2026. NeuroXess, also based in Shanghai, is completing a year-long trial of a fully wireless BCI system and is widely expected to be the next NMPA approval candidate. Neuralink’s CONVOY and VOICE trials are both active. Precision Neuroscience received FDA 510k clearance in April 2025.

The race is no longer theoretical. It is clinical, commercial and geopolitical all at once.

What happened with NEO is only one piece of a much larger story about where brain technology is heading. For a deeper look at how biological neural tissue is already being used to power computing itself, visit Strange Happenings.