The US and China are already locked in a Cold War over AI chips, semiconductors, and electric vehicles. Now that rivalry has moved to the most intimate frontier possible: the human brain. And for the moment, China just lapped the field.
In March 2026, China’s National Medical Products Administration (NMPA) approved NEO, a coin-sized invasive brain-computer interface developed by Shanghai-based Neuracle Technology and researchers at Tsinghua University, for commercial use beyond clinical trials. It is the first invasive BCI product in the world to receive such approval, beating Elon Musk’s Neuralink to the market by a margin that could take years for the US company to close.
This isn’t a lab demo. It’s a real medical product with real patients, real outcomes, and, this is the part that shocked most observers, it’s already being integrated into China’s national health insurance system.
Why NEO Beat Neuralink to the Finish Line
For years, the BCI race has been framed as a battle of headlines. Neuralink’s flashy demonstrations, a monkey playing Pong, a human typing by thought, captured global attention. But while Musk was promising “Jesus-level technology”, Neuracle was quietly running clinical trials with a fundamentally different design philosophy.
The critical difference is surgical depth.
| Feature | NEO (Neuracle) | Neuralink N1 |
|---|---|---|
| Placement | On the dura mater (brain’s protective membrane) | Threads penetrate the cortex |
| Surgery time | ~90 minutes | ~2+ hours with robotic surgeon |
| Electrodes | 8 sensors on the membrane | 1,024+ threads penetrating tissue |
| Risk profile | Lower risk of hemorrhage, glial scarring, and signal degradation | Higher invasiveness, potential for long-term tissue damage |
| Approval status | NMPA commercial (March 2026) | FDA clinical trials only (21 participants) |
Avinash Singh, a BCI researcher at the University of Technology Sydney, told MIT Technology Review that NEO’s “relatively less invasive” design is the primary reason for its fast approval. By sitting on the dura mater rather than piercing the cortex, the device avoids the most dangerous regulatory pitfalls, namely, the risk of hemorrhage and the body’s long-term immune response to embedded foreign objects.
Neuralink’s N1, by contrast, requires a custom-built robotic surgeon to thread microscopic electrodes directly into brain tissue. That approach promises higher signal fidelity, but it demands years of safety data that US regulators are understandably reluctant to fast-track.
The Patient Story That Changes Everything
Numbers and specs only tell part of the story. The real power of NEO becomes clear when you look at what it actually did for Dong Hui.
Dong, 39, was paralyzed from the neck down after a car accident in 2018. In November 2024, he became one of the first people in China to receive an invasive BCI through brain surgery. He had signed up for a clinical trial just one month after seeing on TV how another Chinese patient had apparently used a BCI to hold his granddaughter.
The surgery took just over 90 minutes. A week later, Dong started rehabilitation.
“On the ninth day of my training, my right hand successfully grabbed a ball without the glove”, Dong told MIT Technology Review. “That was a miraculous moment.”
After 11 months of daily 2.5-hour training sessions, Dong was able to write his name, a short “Thank you”, and the date. “I couldn’t believe I was able to write again”, he said. “I was so excited I even missed a stroke in my name.”
This is the kind of real-world outcome that regulators care about. Neuracle conducted 36 clinical procedures in total, 4 feasibility trials and 32 multi-center GCP trials, most completed in 2025. The data was sufficient to clear the device for patients aged 18 to 60 with limb paralysis from spinal cord injuries who retain some residual arm function.
Insurance Coverage: The Silent Accelerator
Here’s where the story gets genuinely disruptive for the global BCI industry.
Days after NEO was approved, China assigned it a unique code within its national health insurance system, the first formal step toward partial reimbursement. This is not a small detail. It represents a fundamental difference in how China and the US approach medical technology commercialization.
As Meicen Sun, an information scientist at the University of Illinois Urbana-Champaign, explained to MIT Technology Review: “Being exceptional and being accessible are two diametrically opposed definitions of winning.”
The US definition of winning is about being first to achieve state-of-the-art performance. China’s definition is about capturing consumers and delivering solutions at societal scale. One pursues peak performance, the other pursues volume manufacturing and universal access.
The investor reaction has been telling. Venture capital firm Future Market Insights projects the brain-implant industry will grow to a $1.7 billion market by 2035. But those projections were made before insurance coverage entered the equation. With state-backed reimbursement, the addressable market in China alone could accelerate far faster than any analyst predicted.
The Engineering Reality Behind the Hype
Let’s be precise about what NEO actually does, and doesn’t do.
The device is not a general-purpose brain interface for typing essays or controlling every device in a room. It’s a focused motor-function rehabilitation tool. The system reads aggregate brain signals from the motor cortex through its eight epidural sensors, transmits them wirelessly to a nearby processing hub, and translates those signals into commands for a soft robotic glove.
The approved label is specific:
NEO is intended for patients aged 18, 60 with paralysis in all limbs due to spinal cord injury who retain some residual function in their arms.
This is hospital-grade medical equipment, not a consumer electronics accessory. Patients still need months of rehabilitation, clinical support, and regular follow-up. A brain implant is not just another wearable.
But that precision is exactly what makes the approval so significant. The first commercial brain implant does not need to be the most spectacular one. It needs to solve a painful problem, reduce enough risk for regulators, and fit into an existing medical system. That is how difficult technologies usually enter the world.
The Bigger Picture: China’s BCI Industrial Strategy
NEO’s approval is not an isolated event. It’s the opening salvo in a coordinated national campaign to dominate brain-computer interface technology.
China’s latest five-year plan, published on the same day Neuracle received its approval, lists BCI as one of six key industries for national competitiveness, alongside quantum technology, humanoid robots, and other priority sectors. The government has already begun clustering research labs, hospitals, startups, and manufacturers in Shanghai and other regions.
Beinao-1
Developed by the Chinese Institute for Brain Research in Beijing and its affiliate NeuCyber NeuroTech, targets patients with movement and speech difficulties from spinal cord injuries or ALS. Experts predict approval as early as 2028.
NeuroXess
Has conducted early trials combining BCIs with AI, including a large language model capable of decoding Mandarin speech signals in real time at up to 300 characters per minute, potentially restoring speech to patients who have lost the ability to speak.
StairMed
And other startups are already in field trials with their own approaches.
“There is no comparable national-level ambition or coordinated map elsewhere in the world at the moment”, Singh told MIT Technology Review.
The Ethical and Security Questions Everyone Is Ignoring
Of course, the commercial approval of invasive brain implants opens a Pandora’s box that technology optimists would prefer to keep closed.
Cybersecurity experts are raising uncomfortable questions about cognitive freedom and neural data privacy. Unlike smartphones or smart speakers, which track your location and voice, brain chips intercept your most intimate, unspoken thoughts. This data is a goldmine for surveillance advertising corporations and a dream asset for regimes interested in thought compliance.
Dr. David Tuffley, a cybersecurity expert at Griffith University, warned that compromised brain implants could theoretically allow hackers to “access sensitive neural data, such as patients’ thoughts and memories”, and even “impair a patient’s cognitive functions, such as the ability to concentrate, or manipulate motor signals to affect how well they move.”
A 2025 survey by the Pew Research Center found that while 72% of American adults believe brain-computer interfaces should be regulated like medical devices, only 38% trust that companies developing these technologies will prioritize safety over profits. In China, the trust dynamic is inverted: state-backed approval carries credibility for patients, but the lack of independent oversight raises its own red flags.
The prevailing sentiment on developer forums is that the security challenges of BCI technology are being systematically underestimated. Most discussions focus on surgical risks and clinical outcomes, while the long-term data sovereignty issues remain largely unaddressed.
What This Means for the Global BCI Industry
The most important takeaway is not that China “won” some imaginary race against Neuralink. It’s that China has created a working commercial pathway for an invasive medical device that no other country has yet figured out.
Neuralink remains in FDA-supervised trials with roughly 21 participants. Synchron, which takes a less invasive route through blood vessels, is also in trials. Neither has a US commercial approval date on the calendar.
This doesn’t mean Neuralink’s approach is wrong. Higher signal fidelity from cortical penetration could eventually enable capabilities that NEO’s epidural design cannot match, direct neural read/write, broader device control, and potentially even sensory restoration. Musk has discussed the possibility of healthy humans receiving implants by 2030.
But for now, the learning advantage belongs to China. Every patient implanted with NEO generates real-world data that feeds back into device improvements, surgical protocols, and rehabilitation software. The companies that learn fastest from patients, surgeons, and regulators will shape the next generation of devices.
The Developer Angle: A New Regulated Hardware Category
For software engineers and AI developers, NEO’s approval signals something concrete: BCIs are transitioning from academic research into a genuine regulated hardware category that demands serious technical expertise.
- Low-latency signal decoding, real-time translation of neural signals into digital commands
- On-device inference, ML models running at the edge for immediate response
- Rehab gamification, building engaging training protocols that patients actually want to use
- HIPAA-grade data pipelines, because neural data is the most sensitive personal information imaginable
This is the same talent pool needed for embodied AI, humanoid robots, and physical AI systems. The skills transfer directly.
China’s NEO approval shifts BCIs from a contest of demos into a contest of hospitals, regulators, reimbursement, and manufacturing. That is harder work than grabbing headlines. But it’s how you actually change the world.
For now, Neuracle has given China the first move. The rest of the world is deciding how to respond.
