Tetiana Aleksandrova claims that she began posing the question, “Would you personally get a BCI installed?” at neuroscience conferences. Almost everyone responded in the same way. It must have been much less intrusive. Not in the future. Right now. Repeated in research labs and hospital hallways, this response reveals a crucial aspect of where this technology has been stuck and why the current push to remove it from operating rooms feels like more than incremental advancement.
For many years, brain-computer interfaces—systems that establish direct communication between brain activity and external devices—have been used in research settings. However, surgery is still a part of the version that most people imagine. Electrodes must be inserted into brain tissue using a robotic process for Neuralink’s implant. A permanent implant remains inside the body even with Synchron’s more cautious method, which involves passing a device through the jugular vein instead of breaking the skull. Paralyzed patients have been able to move cursors and send text messages thanks to these serious interventions that are suitable for serious conditions. The outcomes are truly amazing. However, there is a huge barrier to entry.
Selling the non-invasive route has proven more difficult, in part due to the weaker signals. Electrical activity is absorbed and dispersed by the skull before it reaches a sensor when electrodes are placed on the scalp rather than directly on neural tissue. For many years, the primary argument against EEG-based BCIs as anything more than a scientific curiosity was the signal degradation. Artificial intelligence has been the counterargument, evolving first slowly and then quite rapidly. Under controlled circumstances, transformer-based decoders trained on thousands of hours of neural recordings can now extract meaningful information from signals that previously appeared to be noise with accuracy rates close to 95%. Without opening anyone’s eyes, that represents a substantial change in what is truly feasible.
In April 2025, researchers at Georgia Tech made a significant announcement. A group under the direction of Professor W. Hong Yeo created a microneedle sensor that is less than one millimeter in diameter, flexible, wireless, and able to record high-fidelity neural signals for up to twelve hours straight. It is small enough to fit between hair follicles and slightly beneath the skin. The device recorded brain activity while subjects stood, walked, and ran with 96.4% accuracy in tests involving six participants. They made and received AR video calls without using their hands. Gels, large headsets, and obvious hardware were absent. It’s possible that this is how everyday BCI will eventually appear—nearly undetectable, worn like nothing at all.
Meanwhile, a startup called Subsense is creating a BCI that uses a nasal spray to deliver nanoparticles that function as signal receivers inside the brain in what seems to be the most bizarre entry in this field. No electrodes, whether implanted or not. It’s still genuinely unclear if the strategy will make it through clinical trials, but the fact that anyone is trying it speaks to the direction of collective energy in this field.
As a result, the market is shifting. Businesses like Emotiv, Neurable, and OpenBCI are incorporating EEG into headphones, earbuds, and mixed-reality headsets in an effort to reach consumer and business markets that surgical systems are unable to. The MW75 headphones from Neurable, created in collaboration with Master and Dynamic and acknowledged at CES 2026, incorporate brain sensing into a product that people already wear on a daily basis. EEG-based systems account for nearly 60% of the technology share in the $1.33 billion global neurotechnology BCI market, which is expanding steadily.
With its operating rooms, surgical robots, and meticulously controlled patient reveals, it’s difficult to ignore how different this appears from the Neuralink press cycle. Researchers in Atlanta are fitting sensors between hair strands, a Silicon Valley founder is proposing brain interfaces delivered through the nose, and a Boston company is integrating neural decoding into common headphones. The non-invasive side of this field is quieter and, in some ways, more fascinating. No one had to drill a hole to get there, and the gap between the brain and the machine is getting smaller.
