No brain-computer interface is commercially approved anywhere in 2026; every implant, from Neuralink to Synchron, still runs under research protocols. Yet a paralysed man named Gert-Jan Oskam has already walked using intention alone, decoded straight from his brain. Both facts are true at once, and the gap between them is the real story.
What a Brain Chip Actually Does (It's Not Mind Reading)
A brain-computer interface doesn't read thoughts in general. It records electrical activity from a specific patch of motor cortex, the brain region that plans movement, and uses a decoding algorithm trained on that person's own neural patterns to translate "intent to move a cursor left" into an actual cursor movement. The chip captures a narrow, specific signal, not a general stream of consciousness.
That distinction matters because it explains the current ceiling. In January 2024, Noland Arbaugh became the first human to receive Neuralink's N1 implant and was reportedly playing online chess using only decoded intent within weeks. That's real, but it's decoded intent controlling a screen, not the brain directly reactivating a paralysed arm.
The Technology That Actually Restores Movement, Not Just Control
Restoring a person's own limb movement requires a second piece: a brain-spine interface. Neuroscientist Grégoire Courtine and colleagues at NeuroRestore in Lausanne built a system that reads motor-cortex signals and uses them to drive epidural electrical stimulation directly on the spinal cord, below the injury site, reactivating the muscles those signals were always meant to control.
Gert-Jan Oskam, paralysed in a cycling accident, used this system to walk again, and researchers reported residual improvements in his voluntary movement that persisted even after the interface was switched off, consistent with the nervous system rebuilding some of its own connections. That single mechanism, cortex-to-spine rather than cortex-to-cursor, is the difference between "controlling a device" and "moving your own leg."
Three Companies, Three Very Different Bets
Neuralink's N1 implant uses thousands of ultra-thin electrode threads inserted directly into cortical tissue for high-resolution signal, but early implants saw some threads retract from their optimal position over time, a hardware problem the company says its revised second-generation design addresses. Synchron takes the opposite trade-off: its Stentrode device is threaded through blood vessels with no open-skull surgery at all, using just 16 electrodes, and its COMMAND trial reported zero serious adverse events across implanted patients with 100% successful device deployment.
Precision Neuroscience sits between the two, placing a 1,024-electrode array on the brain's surface through a minimally invasive slit rather than a full craniotomy, and had filed for what may be the field's first premarket approval application by 2025. None of the three has converted early results into a commercially available product yet.
Why "Available in 2026" Is the Wrong Headline
As of 2026, no brain-computer interface for paralysis has full FDA premarket approval. Every device covered here operates under research trials or expanded access programs, and realistic timelines for limited commercial availability run through 2028 to 2030. That's a specific, checkable claim, not caution for its own sake, and it separates responsible coverage from the promotional framing that dominates most search results on this topic.
The distinction has consequences beyond marketing accuracy. Patients researching enrollment, and families weighing cost, are better served by knowing that pricing, insurance coverage, and long-term device reliability remain unresolved questions rather than settled ones. This is the same investigational-versus-market-ready gap that shows up whenever CRISPR-based gene editing therapies get covered before their own regulatory milestones are reached.
The Part Neither Optimists Nor Skeptics Get Fully Right
Coverage of BCIs tends to split into two camps: breathless claims that paralysis is essentially solved, and dismissive framing that treats every implant as a lab curiosity. Neither matches the 2026 evidence. Motor restoration through brain-spine interfaces is real and published in peer-reviewed journals, but it remains a small-cohort, early-phase result, not a standard treatment available to the roughly 5.4 million people living with paralysis in the United States alone.
The honest middle position, and the one this article is built around, is that the mechanism works in principle and has already worked for at least one named patient, while the infrastructure to deliver it broadly, safely, and affordably is still years away. That gap between demonstrated science and delivered medicine is worth tracking the same way protein-folding prediction tools took years to move from breakthrough paper to practical drug pipeline.
What to Actually Watch For Next
The next real milestone isn't a flashier demo video, it's a completed FDA pathway. Watch Synchron's protocol publication, Neuralink's 30-patient PRIME trial completion, and whether Precision Neuroscience's premarket submission converts into an approved device rather than another expanded-access program. Whichever company clears that bar first will likely define the regulatory template the rest of the field follows for years afterward.
Frequently Asked Questions
Can a brain chip already restore movement to paralysed patients?
Yes, in early clinical research: a brain-spine interface built by Grégoire Courtine's team at NeuroRestore let a paralysed man named Gert-Jan Oskam walk by routing decoded brain signals to spinal cord stimulation. It is not yet a commercially available treatment.
Is any brain-computer interface FDA approved in 2026?
No. As of 2026, every BCI for paralysis, including devices from Neuralink, Synchron, and Precision Neuroscience, operates under research trials or expanded access programs, with commercial approval realistically expected between 2028 and 2030.
What's the difference between a brain-computer interface and a brain-spine interface?
A brain-computer interface typically decodes intent to control an external device like a cursor or robotic arm. A brain-spine interface uses that same decoded signal to drive electrical stimulation on the person's own spinal cord, reactivating their own muscles.
Which company's brain chip is safest?
Synchron's Stentrode, which is threaded through blood vessels without open-skull surgery, reported zero serious adverse events across implanted patients in its COMMAND trial as of late 2024, trading electrode count and resolution for a less invasive procedure.
How many electrodes do these brain implants use?
Synchron's Stentrode uses 16 electrodes. Precision Neuroscience's Layer 7 array uses 1,024. Neuralink's N1 implant uses thousands of individual threads, prioritizing signal resolution over surgical invasiveness.
Will brain chips eventually treat conditions beyond paralysis?
Researchers are already testing related neurotechnology for treatment-resistant depression using closed-loop deep brain stimulation, but that application is a separate, earlier-stage research track from the motor-restoration work covered here.