The device is called Neo, and for the first time, technology of this kind has moved beyond the confines of clinical trial participants to receive authorization for use by patients meeting established medical criteria. Approval was granted in March of this year by Chinese regulators to Neuroco Medical Technology, a Shanghai-based company.

According to *Nature Biotechnology*, this marks the world's first commercial approval for an invasive brain-computer interface; however, there is a key difference compared to Neuralink: China opted not to place the device quite so close to the neurons. Neo is roughly the size of a coin and uses only eight electrodes; rather than inserting dozens of filaments directly into brain tissue, the device is placed atop the dura mater—one of the membranes protecting the brain—over the region of the cortex responsible for controlling movement.

This means the signal captured is less detailed than that obtained by systems that penetrate the tissue directly, but there is an advantage: the brain does not need to be pierced by electrodes. The concept behind Neo is relatively simple: a paralyzed person imagines moving their hand; even though a spinal cord injury prevents the signal from reaching the muscles, the brain continues to generate activity associated with the intention to move.

The eight electrodes capture these signals and send them to a computer; an algorithm identifies the intent and transmits the command to a pneumatic robotic glove worn on the patient's hand—so when the person thinks about moving their hand, the glove executes the movement. To secure approval, the system underwent clinical validation involving 32 patients, conducted in collaboration with 11 Chinese medical institutions; another interesting detail is that the implant lacks a conventional battery requiring surgical removal for replacement. A coil positioned outside the head transmits energy via a magnetic field to another coil within the implant. That same connection allows brain signals to be transmitted outward.

The team sought to eliminate both the implanted battery and the wires permanently passing through the skin—two elements that could cause complications in a device designed to remain in the body for years. It was precisely this approach that allowed China to reach a milestone Neuralink has not yet achieved: moving beyond the experimental phase and entering the market. However, this does not mean just anyone in China can walk into a hospital and request a brain chip; the approved indication for use is quite specific.

The NEO device is intended for individuals aged 18 to 60 suffering from paralysis of all four limbs due to specific cervical spinal cord injuries. Yet, NEO is only part of the broader picture, as China is also developing another system called "Bay Now One." Here, the scale of operations is expanding; in July, Chinese scientific authorities reported that the system had already been used in 16 implants. The participant with the longest-standing implant had already surpassed the one-year mark; collectively, the systems had logged over 55,000 hours of operation, and participants had used the interface to control robotic arms with their thoughts and work on restoring motor functions.

The announced plan aims to reach 36 implants by 2026 and launch trials at top-tier hospitals across the country in 2027. This reveals that the story extends far beyond a mere race between Neuralink and a Chinese company. The Chinese government has designated brain-computer interfaces as one of the "industries of the future," while companies, hospitals, and research centers simultaneously advance various technologies in the field.

So, we are beginning to see two strategies evolving in parallel: in the United States, Neuralink is attempting to build a high-bandwidth interface capable of transforming brain activity into increasingly complex commands—and now even into speech—while in China, various groups are exploring approaches that seek to balance capacity, safety, and clinical implementation. Yet, both are essentially trying to solve the same problem: creating a direct channel between the human brain and a machine.

In my view, this is precisely where the race becomes truly fascinating to watch. Until now, we have mostly discussed people controlling something external to the body—a cursor, a glove, a robotic arm, or a synthetic voice—but researchers are already working in the opposite direction. Instead of asking only how to extract information from the brain, they have begun to ask how we can send information directly into it. And as that door begins to open, the brain-computer interface ceases to be merely a way to control machines; it can start to transform into a means of altering what a person perceives.