Brain-Computer Interface 2026: 7.2 Billion Capital Influx, Industrialization Is Crossing the Critical Point
On July 1, Elon Musk's Neuralink announced the completion of the world's first dura-mater-penetrating brain-computer interface implantation surgery. The step of cutting the dura mater was completely removed, and electrode wires were directly penetrated through the dura mater and implanted into the brain itself, so as to reduce surgical trauma, shorten recovery time, and create conditions for larger-scale clinical applications in the future. This latest development has once again brought brain-computer interface into the spotlight of the global technology and capital markets.
In China, the brain-computer interface track is also entering a period of capital boom, with financing in the primary market accelerating in an all-round way. Earlier this year, the news of BrainCo's 2 billion yuan financing kicked off the investment boom in the track, and the official release of the world's first brain-controlled robot training platform preview a few days ago also attracted widespread attention in the industry. According to data from ITjuzi, as of June 30, the investment and financing transactions in the brain-computer interface track reached 64, with a total amount of about 7.253 billion yuan, far exceeding the total transaction volume of 2025.
Behind the boom is the epoch-making significance represented by the brain-computer interface technology itself.
Nowadays, brain-computer interfaces are building pathways between the human brain and silicon-based peripherals. In the future, activities that rely on humans to perceive the environment, make judgments and perform operations may find brand-new implementation methods as a result. It is this versatility that crosses physical constraints that makes it no longer limited to a single segmented field, but explores landing scenarios in medical care, consumption, human-computer interaction, games and entertainment and other fields.
In medical rehabilitation, brain-computer interface systems are helping patients re-establish motor functions; in the field of education, it can capture and adjust attention status in real time; in consumer electronics, human-computer interaction and other aspects, brain-computer interfaces are also bringing more immersive experiences.
According to the degree of invasiveness of the device to the brain, brain-computer interfaces can be divided into three categories: non-invasive, semi-invasive and invasive. Neuralink founded by Elon Musk and Merge Labs invested and founded by Sam Altman have bet on the invasive and non-invasive routes respectively, and have also become important vane for investment and financing in this track.
In China, the brain-computer interface industry presents a trend of parallel development of three routes, each anchored in different application scenarios, corresponding to different commercialization rhythms and investment logics. The invasive route features high risk and high return, which is more suitable for capital with long-cycle layout; the non-invasive route has realized commercial landing, with a clear large-scale revenue path, attracting diverse industrial capital with low risk and stable return; the semi-invasive route is between the two.
Technical Divide Between Invasive and Non-Invasive Routes
With the advantage of directly contacting the cerebral cortex, invasive brain-computer interfaces have high signal accuracy and signal-to-noise ratio, which is the core technical path to overcome severe neurological diseases. Its core research and development directions focus on implantation trauma control and long-term stability in vivo. Of course, this route also corresponds to stricter medical device regulatory requirements and longer clinical verification cycles.
The core challenge of non-invasive routes lies in the "extracranial extreme signal-to-noise ratio": the skull attenuates more than 80% of neural electrical signals, and the collected signals are often weak signals at the microvolt level, mixed with a large amount of interference. To accurately decode human motion intentions, emotional states and even thinking contents from noisy signals, the engineering difficulty of signal processing and algorithm decoding is no less than the accuracy challenge of surgical implantation — which is equivalent to listening to the whisper on the other side of a wall through the wall, and distinguishing every single word.
This is also the reason why dry electrodes and deep learning decoding algorithms have become the core research and development directions of non-invasive routes. Represented by BrainCo, the company's self-developed solid-state gel electrodes break through the large-scale bottleneck of traditional wet electrodes that require conductive paste smearing and cumbersome wearing. Combined with low-noise circuit design, the signal-to-noise ratio of real EEG test reaches 10.78dB. On this basis, its high-performance neural signal decoding algorithm can integrate multi-modal physiological signals, complete the transformation from intention to action instruction within 200 milliseconds, and its motion intention recognition accuracy is at the leading level in the industry.
Around the world, the upper accuracy limit of non-invasive brain-computer interfaces is also being continuously refreshed. In June 2026, Meta and the Basque Cognitive Center of Spain jointly released the Brain2Qwerty v2 model, which increased the decoding accuracy of non-invasive brain-computer interfaces from 8% to 61%, close to the level of some invasive solutions. In the field of tactile feedback, BrainCo has realized a two-way closed loop from environmental perception to neural feedback, identifying object features through flexible tactile sensors, and then converting contact information into micro-stimuli perceptible to the human body, allowing users to perceive subtle environmental changes through neural and myoelectric pathways, improving the immersion and operation accuracy of human-computer interaction.
Simultaneous Landing in Medical and Consumption Fields: Commercial Prospects of Non-Invasive Brain-Computer Interfaces
The underlying logic of all countries in the world supporting the brain-computer interface industry is highly consistent: encourage technological innovation, but emphasize safety, accessibility and benefits for the public.
The relevant research of the US NIH "BRAIN Initiative" and DARPA has always supported the parallel development of multiple routes including non-invasive routes; the European Union has put forward strict requirements on the safety and ethics of brain-computer interfaces in the Human Brain Project and AI regulatory framework, giving priority to encouraging technical solutions with controllable risks and wide application prospects.
At the same time, domestic policies are also paving a clear path for the industrialization of brain-computer interfaces.
In July 2025, seven departments including the Ministry of Industry and Information Technology of China jointly issued the *Implementation Opinions on Promoting the Innovation and Development of the Brain-Computer Interface Industry*, which clarified two-phase goals, proposed to accelerate the application of brain-computer interface products in industrial manufacturing, medical health, life consumption and other fields by 2027. The core orientation is to take scenario landing as the traction, so that the technology can truly serve real demands.
Among them, non-invasive technology, with the features of safety, non-invasion and short landing cycle, has been applied in multiple segmented scenarios. In terms of product qualifications, products such as BrainCo's intelligent bionic prostheses and ADHD rehabilitation training systems have successively obtained authoritative certifications including FDA and NMPA, covering multiple clinical demands such as limb function reconstruction and neurodevelopmental disorders. From the payment side, the National Healthcare Security Administration has set up an independent charging category for non-invasive brain-computer interfaces. Zhejiang, Beijing and other regions have included it in the medical insurance reimbursement scope, and its commercial sustainability has been continuously strengthened.
Apart from medical rehabilitation, brain-computer interfaces also have market potential on the consumer side. As early as around 2010, Muse launched a neurofeedback headband focusing on sleep and meditation training, and Emotiv's EPOC opened up another market in game development and human-computer interaction research.
Recently, the application capability of brain-computer interfaces in human-computer interaction has also continued to break through. According to the announcement of BrainCo, the 21-high-degree-of-freedom dexterous bionic hand has been iteratively applied. At the upcoming Shanghai World Artificial Intelligence Conference, its cutting-edge achievements including the first brain-controlled robot system and neural data acquisition solution will also be unveiled, providing underlying support for the development of embodied intelligence from both ends of signal decoding and data supply respectively.
Combined, these advances point to a more accessible future. Brain-computer interfaces are stepping out of laboratories and penetrating into diverse scenarios, assisting in improving concentration in education, creating immersive experiences in entertainment, providing precise health interventions on the consumer side, and building a neural interaction base in the field of embodied intelligence.
The ultimate value of truly epoch-making technologies lies in their "popularization". So is electricity, so is the Internet, so is brain-computer interface. When a technology can only serve a very small number of people, it is only a cutting-edge exploration in the laboratory; when it can enter the public life with affordable cost and convenient ways, it truly opens the era of the whole industry. On the long road of brain-computer interface from technology to universal benefit, the non-invasive route is taking the lead in entering the critical transition period of industrialization.