China has approved the world’s first invasive brain-computer chip—here’s what’s next

In October 2024, a 39-year-old man named Dong Hui sat in the courtyard of his home in Henan province, China, and accomplished what had seemed impossible six years earlier following a catastrophic spinal cord injury from a car accident. Paralyzed from the neck down, Dong held a pen and wrote his own name, the words "Thank you," and the date—a feat that would have been unimaginable without the invasive brain-computer interface surgically implanted in his skull. This breakthrough moment preceded an even more significant milestone: in March 2025, the device Dong uses, called NEO and developed by Shanghai-based startup Neuracle Technology in collaboration with Tsinghua University in Beijing, became the world's first invasive brain-computer interface approved for clinical use outside experimental trials. The approval, granted by China's National Medical Products Administration, represents a watershed moment in neurotechnology, transforming what was once purely experimental intervention into a regulated medical treatment available to patients with specific paralysis conditions caused by spinal cord injury.

The development of invasive brain-computer interfaces has occupied neuroscientists and biomedical engineers for decades, with early research dating back to the 1990s when scientists first demonstrated the feasibility of decoding motor intentions from neural signals recorded directly from the brain. What distinguishes the current moment from previous decades of incremental progress is the convergence of multiple enabling factors: miniaturization of electrode arrays, advances in machine learning algorithms that decode neural signals with greater accuracy, improvements in biocompatible materials that allow long-term implant stability, and critically, regulatory frameworks willing to approve these devices for broader patient populations. NEO's approval arrives at a time when spinal cord injury represents a persistent challenge in global healthcare, with limited therapeutic options available to restore functional movement in severely paralyzed individuals. China's decision to approve this technology reflects both the maturation of the underlying science and a strategic positioning within the global competition for neurotechnology leadership, particularly significant given that rival systems from established players such as Neuralink, the California-based company founded by Elon Musk, have not yet achieved equivalent regulatory clearance outside their home jurisdictions.

The technical specifications of NEO reveal careful engineering choices that have enabled its clinical success. The device comprises a coin-sized sensor array implanted subdurally—placed on the dura mater, the tough protective membrane surrounding the brain tissue itself—rather than directly within the brain parenchyma, which reduces invasiveness while maintaining signal quality. During Dong's implantation procedure, which lasted slightly over ninety minutes, the electrode array was positioned to capture motor cortex signals, which were then transmitted wirelessly to an external processing unit via a transdermal connector implanted on his skull. The computational processing translates raw neural signals into movement commands for a soft robotic glove that Dong wears during daily two-and-a-half-hour training sessions. The regulatory approval encompasses patients aged 18 to 60 with complete paralysis in all limbs due to spinal cord injury but retaining some residual arm function—a carefully defined patient population that represents a meaningful but circumscribed segment of individuals living with severe paralysis. Dong's clinical trajectory exemplified the device's efficacy: beginning rehabilitation approximately one week after surgery, by his ninth training session his right hand successfully grasped a ball without glove assistance, demonstrating direct neural control of his own muscles rather than merely robotic manipulation.

For individuals living with severe spinal cord injuries, NEO's approval signifies something distinctly different from previous neurotechnology announcements: a device one can actually access through standard medical channels rather than through experimental protocols. Dong's expressed motivation—to regain sufficient hand control to dress himself, eat independently, and reduce dependence on his aging parents—encapsulates the tangible stakes of this technology. The functional improvements that emerge from brain-computer interface training are not merely symbolic; they represent restoration of genuine agency in activities of daily living that most people perform without conscious deliberation. The regulatory pathway that NEO traversed in China establishes crucial precedent and baseline evidence that will influence how other nations' regulatory bodies evaluate comparable technologies. Moreover, the approval signals to potential patients within China's substantial population of individuals with traumatic spinal cord injuries that they now possess a concrete therapeutic option rather than merely an experimental possibility. This transition from research subject to patient represents a fundamental shift in how neurotechnology moves through the innovation pipeline, establishing economic viability for companies developing such systems and demonstrating that commercial-scale manufacturing, quality assurance, and patient support infrastructure can sustain these complex interventions.

The broader significance of NEO's approval extends beyond individual patient outcomes to illuminate fundamental questions about technological development pathways and regulatory innovation. China's willingness to approve an invasive neural implant reflects a regulatory approach that emphasizes clinical benefit for patients with limited alternative treatments, even when long-term safety data remains necessarily incomplete—a calculus that differs from regulatory frameworks in some Western jurisdictions that may require extended longitudinal data collection before approval. NEO's achievement of regulatory clearance before competing systems from better-capitalized companies suggests that regulatory speed and approval pathways matter as much as raw research resources in determining which technologies reach patients first. Furthermore, the concentration of brain-computer interface development in both China and the United States represents a geopolitical dimension of neurotechnology, with implications for intellectual property, standards development, and which regions capture commercial advantage in what may become a substantial medical device market. The NEO approval demonstrates that invasive brain-computer interfaces are transitioning from speculative future technology to implemented clinical practice, raising accompanying questions about equitable access, training requirements for surgical implantation, liability frameworks, and social integration of individuals whose neural activity is being continuously recorded and interpreted by external devices.

Observers of the neurotechnology landscape should monitor several developments that will clarify whether NEO's approval represents an inflection point or a narrow precedent applicable only to China's regulatory environment. Neuralink's progression through United States regulatory processes and clinical trial outcomes in human subjects will provide crucial comparative data regarding alternative design approaches and approval timelines in Western regulatory contexts; the company has conducted initial implantations but has not yet achieved comparable regulatory approval outside experimental frameworks. Neuracle Technology's expansion plans and manufacturing capacity will reveal whether the company can sustainably serve growing patient populations or whether supply constraints limit access. Additionally, outcomes tracking from NEO's broader clinical deployment beyond initial trial participants will generate real-world evidence regarding durability, infection rates, revision surgery requirements, and sustained functional improvements, data that will either validate or constrain the regulatory precedent China has established. By 2026, the landscape of approved invasive brain-computer interfaces globally will likely become substantially clearer, establishing whether China's regulatory pathway accelerates broader adoption or remains a distinctive regulatory choice with limited replication elsewhere. The fundamental question now centers not on whether invasive neural interfaces work in principle, but rather which regulatory frameworks, commercial models, and clinical support systems can sustain them as standard medical practice.