Brain-Computer Interfaces Enter the Commercial Era as China Grants First Approval and BrainGate Achieves Near-Able-Bodied Typing Speeds

Overview

In the span of a single week, two developments have shifted brain-computer interfaces from laboratory curiosities to practical clinical tools. On March 13, China’s National Medical Products Administration (NMPA) granted the world’s first commercial marketing approval to an implantable brain-computer interface (BCI), a device built by Shanghai-based Neuracle Medical Technology. Three days later, researchers from Mass General Brigham and Brown University published results in Nature Neuroscience showing that their BrainGate typing neuroprosthesis enabled a paralyzed participant to type at 110 characters per minute — roughly 80 percent of able-bodied texting speed — with a word error rate of just 1.6 percent.

Taken together, the milestones mark the moment BCIs crossed from experimental research into two distinct but complementary frontiers: regulatory commercialization and high-performance communication restoration.

China’s Commercial Breakthrough

Neuracle’s device is a coin-sized wireless implant that sits on the surface of the brain’s outer membrane rather than penetrating cortical tissue, an approach that reduces the risk of neural damage compared to penetrating electrode arrays. The system reads electrical signals from the motor cortex and decodes them to control a robotic glove, restoring grasping function for patients with cervical spinal cord injuries.

The NMPA approval followed a clinical program that included 36 implant procedures — four feasibility cases and 32 multicenter confirmatory cases — with endpoint collection completed by late November 2025, according to Scientific American. All patients achieved home-based brain-controlled grasp assistance and rehabilitation training, and no device-related serious adverse events were reported. The approval covers adults aged 18 to 60 with chronic cervical spinal cord injury who retain some upper-arm function but cannot grasp.

The significance extends beyond the device itself. As Scientific American noted, this marks the first time any nation has approved an invasive BCI for broad patient use outside clinical trials. Doctors can now prescribe and implant the technology as they would a pacemaker or deep brain stimulator — a regulatory precedent that no Western device has yet achieved.

Neuracle’s stock surged up to 10 percent following the announcement, according to Interesting Engineering. The approval aligns with Beijing’s designation of BCI as one of six strategic industries of the future in its latest five-year plan, with the government pledging to streamline regulatory review and establish reimbursement guidelines, as reported by the South China Morning Post.

BrainGate’s Performance Leap

On March 16, the BrainGate consortium published results demonstrating what may be the most practical BCI communication system to date. The typing neuroprosthesis uses microelectrode sensors implanted in the motor cortex to detect attempted finger movements. A QWERTY keyboard is displayed with each letter mapped to specific finger positions — up, down, or curled — and as participants intuitively attempt these movements, the system translates the neural activity into typed characters, according to the study published in Nature Neuroscience.

Two participants — one with advanced amyotrophic lateral sclerosis (ALS) and one with cervical spinal cord injury — operated the system from their homes. One reached 110 characters per minute, equivalent to 22 words per minute, with a word error rate of 1.6 percent that matches able-bodied typing accuracy. Calibration required as few as 30 sentences.

“BCIs are on track to become an important new alternative to what’s currently offered,” said senior author Dr. Daniel Rubin of Mass General Brigham, as reported by Brown University. The remark reflects a specific frustration: existing augmentative communication systems, particularly eye-gaze technology, are often slow and fatiguing for patients with severe motor impairments.

First author Dr. Justin Jude highlighted that decoding finger movements opens a broader pathway, as the approach could enable “complex reach and grasp movements” and support personalized keyboard interfaces, according to STAT.

The Competitive Landscape

The two breakthroughs arrive amid intensifying global competition in the BCI sector. Neuralink, which has implanted approximately 20 patients with severe paralysis, plans to expand trials to Canada, the United Kingdom, Germany, and the United Arab Emirates in 2026 but has not yet received commercial approval from any regulatory body, as Scientific American noted. Synchron, whose stentrode device threads through blood vessels to avoid open brain surgery, has implanted 10 volunteers across the United States and Australia — the most simultaneous patients reported by any BCI group — and is preparing for a large-scale clinical trial ahead of its own commercial approval bid.

The approaches differ fundamentally. Neuralink’s ultra-high-bandwidth chip threads thousands of micro-electrodes into the cortex using a robotic surgeon, prioritizing signal density. Synchron’s endovascular approach avoids craniotomy entirely, trading bandwidth for lower surgical risk. Neuracle’s surface-mounted design sits between them, resting on the brain’s outer membrane without penetrating tissue. BrainGate’s research-grade system uses penetrating microelectrode arrays for maximum signal fidelity but remains confined to clinical trials.

As STAT reported in December 2025, three trends were expected to define the BCI field in 2026: participation in clinical trials scaling from single digits to dozens of patients, international expansion of trials, and the transition from research tools to commercial products. By mid-March, all three are materializing simultaneously.

What We Don’t Know

Several critical questions remain unanswered. Neuracle’s approval covers a narrow patient population — cervical spinal cord injury with preserved upper-arm function — and no pricing information has been disclosed. The long-term durability of the surface-mounted implant is not yet established; brain implants can shift position or develop scar tissue that degrades signal quality over time. Whether China’s reimbursement guidelines will make the device accessible beyond wealthy urban hospitals remains to be seen.

For BrainGate, the typing system performed impressively but was tested in only two participants, and the microelectrode arrays require invasive surgical implantation. The pathway from research publication to regulatory approval and commercial availability remains long and uncertain. No timeline has been announced for seeking FDA clearance of the typing neuroprosthesis.

Perhaps the largest open question is whether any BCI will achieve the combination of high performance, surgical safety, long-term reliability, and regulatory clearance needed for mainstream adoption. Each current approach excels in some dimensions while accepting trade-offs in others.

Analysis

China’s decision to grant commercial approval carries strategic implications that extend beyond healthcare. By moving first on BCI regulation, Beijing is establishing the standards and clinical infrastructure that could shape the global market. The designation of BCI as a strategic industry, paired with streamlined regulatory review, mirrors China’s approach to electric vehicles and renewable energy — sectors where early government support translated into global market leadership.

For the United States and Europe, China’s approval creates regulatory pressure. The FDA’s approach to BCI devices has been cautious — it initially rejected Neuralink’s trial application in 2022 before approving it the following year. While caution is warranted for devices implanted in the brain, the gap between Chinese commercial availability and American clinical trials could accelerate FDA deliberations on BCI regulatory pathways.

The BrainGate results, meanwhile, demonstrate that the performance ceiling for BCI communication is rising rapidly. A word error rate of 1.6 percent at 22 words per minute begins to approach the threshold where a BCI could serve as a primary communication device rather than a last resort. Combined with the system’s home-use capability and minimal calibration requirements, the technology is converging on practical clinical utility.

The week of March 13-16, 2026, may be remembered as the moment brain-computer interfaces ceased to be a promise and became a product — at least in one country — while simultaneously proving they can perform at levels that justify the surgical risks involved.