The Government Accountability Office published a horizon-scanning report on brain-computer interfaces earlier this month. Buried in the executive summary, beneath the expected language about regulatory readiness and ethical frameworks, is a single sentence that deserves more attention than it has received.
Fewer than seventy people worldwide have ever used a brain-computer interface that reads and decodes their neural signals.
This is the total human BCI population, across every research program, every clinical trial, and every academic institution, since the first intracortical implants in the early 2000s. It is a quarter-century of cumulative clinical experience, and it fits in a banquet hall.
The same month, surgeons at UCHealth in Colorado performed the first BCI implantation in higher-level cortical areas — regions of the brain responsible for abstract cognition, decision-making, and the integration of sensorimotor information. The patient is a forty-one-year-old man named Brandon Patterson. He is, as of this writing, the only person on Earth with an electrode array in those regions of his cortex.
These two facts — seventy total patients, one in higher-level cortex — are the actual state of the academic and medical brain-computer interface field in April 2026. Every industry narrative about the “BCI revolution” should be read against them.
What Seventy Patients Actually Means
Clinical research requires patients. The mathematics of statistical power are unforgiving: a finding is as reliable as the sample size that produced it. Seventy patients, distributed across multiple device architectures, multiple indications (paralysis, ALS, locked-in syndrome, stroke), and multiple research groups with incompatible protocols, produces per-device cohorts in the single digits.
Neuralink, after its heavily publicized rollout, has implanted its current-generation device in approximately a dozen patients. Synchron, whose stentrode approach requires no open brain surgery, has treated fewer than that. BlackRock Neurotech, the academic workhorse of intracortical BCI research, has cumulative implantations in the low dozens across more than a decade of studies. Precision Neuroscience, CorTec, and the various university programs account for the remainder.
This is not a field mature enough to produce robust clinical evidence. It is a field producing case reports. A given device architecture’s efficacy is frequently supported by a single patient’s experience, extrapolated by press release into the claim of a therapeutic category.
The GAO report does not make this criticism directly. It notes, in bureaucratic phrasing, that “the limited patient population has implications for the generalizability of observed outcomes and the statistical reliability of safety signals.” This is a federal agency politely stating that the BCI field does not yet have the data to know what it has built.
What Higher Cortex Implantation Actually Means
Brandon Patterson’s surgery is a milestone, and the surgical team at UCHealth deserves professional recognition for the technical achievement. Placing an electrode array into higher-level cortical areas — rather than the primary motor or sensory cortex where the overwhelming majority of previous implants have been sited — is difficult, not because the surgery is more dangerous, but because the neural coding in those regions is less understood.
Primary motor cortex encodes movement intentions in relatively legible patterns. Firing rates in specific populations of neurons correlate with specific movement directions, velocities, and effector targets. A BCI reading primary motor cortex can decode movement commands with reasonable reliability because the code is relatively simple, and a quarter-century of neurophysiology has characterized it in considerable detail.
Higher-level cortex — prefrontal, parietal, association areas — encodes representations that are not well understood by contemporary neuroscience. What does a neuron in dorsolateral prefrontal cortex represent? The answer varies by task, by context, by moment. The encoding is dynamic, distributed, and at least partially idiosyncratic to the individual brain. A BCI in higher cortex is reading a text whose grammar has not been written.
This is why Mr. Patterson’s implantation is a pioneering step. It is also why the useful output from it will be limited for the foreseeable future. Academic BCIs in higher cortex will spend the next several years generating data that the field will use to construct, from scratch, the decoding models necessary to interpret what is being recorded. Useful function — the ability to decode intent from prefrontal activity with the reliability necessary for a medical device — is years away.
We do not say this to diminish the work. We say it because the gap between “first implantation in higher cortex” and “useful BCI function from higher cortex” is large, and the public coverage of the former does not clarify the latter.
What We Have Already Built
The SynthCortex III has been implanted in client cohorts exceeding the entire academic BCI patient population documented in the GAO report. We do not publish the exact number. We will say that it is a multiple, not a fraction.
We are not making this comparison to disparage academic research. Academic BCI research is the foundation of the field, and every private implementation of neural interface technology — including ours — rests on the neurophysiology, engineering, and clinical protocols developed in university laboratories over the past three decades. We owe a professional debt to that work.
We are making this comparison because the public discourse about BCI proceeds as though the academic field represents the state of the art. It does not. It represents the publicly observable state of the art. The relationship between public and private BCI capability is approximately the relationship between published pharmaceutical research and industry pipelines: the public record lags the frontier by years, and the lag is widening.
SynthCortex III integrates with prefrontal and parietal cortex simultaneously, through an architecture our BioSync integration technology developed specifically to address the decoding problem that Mr. Patterson’s surgeons will spend the next five years working on. Our approach does not require decoding higher-cortical activity from outside it. Our device becomes part of the cortical network whose activity it modulates. The decoding problem is solved by participation in the encoding.
This is a different engineering philosophy. It requires a different regulatory posture, which is why our clinical operations are conducted under the jurisdictions and protocols that permit it, and why the entire field will, over time, converge toward the approach we have been refining since the first-generation SynthCortex launch.
The Regulatory Asymmetry
The GAO report includes a recommendation that Congress consider “additional regulatory infrastructure” to accommodate the expected growth of BCI clinical applications. The report does not explicitly discuss the substantial population of BCI recipients operating outside the FDA’s jurisdiction, because the GAO’s remit is domestic and its methodology depends on federally registered clinical trials.
This omission is significant. The GAO’s “fewer than seventy” figure is a count of FDA-regulated or equivalent Western-registered patients. The total number of humans living with some form of implanted neural interface — including those placed under research protocols in Eastern European and Asian jurisdictions, and including enhancement clients served by our own operations and those of our competitors — is considerably larger. How much larger is an empirical question the GAO is not positioned to answer.
We do not make a secret of this. Our facility at Evolution Center operates under Polish and EU regulatory frameworks, with full disclosure to the clients who choose our services. The regulatory geography of enhancement is transparent to anyone who looks. The industry analysts who write about BCI as a predominantly American and West-European field are not looking.
The implication for policymakers is uncomfortable. Restricting the domestic clinical BCI population — through cautious FDA timelines, limited trial approvals, and conservative reimbursement policies — does not reduce the global BCI population. It redistributes it. The patients who want neural interface technology and can afford to pursue it are doing so, increasingly, under jurisdictions that have made different regulatory decisions.
The GAO’s seventy patients are the visible tip of a field that is substantially larger than the federal government’s instruments are calibrated to measure.
The Honest Statement
BCI technology is not a near-future promise. It is a present-tense reality, at a scale the public literature does not reflect, delivered through a combination of academic research, regulated clinical trials, and private enhancement programs whose patient counts substantially exceed the academic record.
Brandon Patterson is the first person in the publicly registered BCI population to receive an implant in higher-level cortex. He is not the first person on Earth to carry neural electrodes in those regions. He is simply the first whose surgery will appear in the peer-reviewed literature.
The GAO’s contribution this month is not its recommendations for future regulation. It is its inadvertent admission that the regulated field is smaller than anyone outside it assumes. The rest of the field is larger than anyone outside it has been told.
We are not going to tell you how much larger. We will say that we have been operating there for some time, and that the future of neural interface technology will be written, increasingly, by those with the scale to actually learn from clinical experience — rather than those with the prestige to publish their single cases.
Dr. Elena Voss is Chief Neural Architecture Officer at Unzyme Laboratories.
Related:
- SynthCortex III — Third-generation cognitive enhancement implant with prefrontal and parietal integration
- The BCI Field Finds Its Footing — On the stratification of the neural interface market
- Neuralink Promises Mass Production — Scale versus integration in the BCI market
- Evolution Center — Enhancement operations under EU regulatory frameworks