Brain-Computer Interfaces Are Already Working. Here's How Far They've Actually Gotten

Of everything on the frontier of technology right now, brain-computer interfaces might be the strangest to say out loud and mean literally: there are people alive today who control a computer, a phone, or a robotic limb using only their thoughts, via a device implanted in their brain. It's no longer a hypothetical. It's also, honestly, nowhere near the mass-market future either leading company's marketing implies.

What a brain-computer interface actually does

Strip away the sci-fi framing and a BCI does one specific job: it records electrical activity from neurons and uses machine-learning software to translate that activity into a digital command — move a cursor, select a letter, close a hand. The people using these devices today are almost entirely patients with severe paralysis from conditions like ALS or spinal cord injury, for whom the technology restores a basic ability most people never think about: controlling a device without moving a muscle.

Two very different bets on how to get there

Neuralink and Synchron have taken deliberately opposite approaches. Neuralink's N1 implant packs 1,024 electrodes for rich, high-fidelity neural data, but requires a surgical procedure that removes a section of skull and inserts electrode threads directly into brain tissue. Synchron's Stentrode takes the opposite trade: a 16-electrode array shaped like a stent, delivered through a blood vessel via catheter and lodged in a vein next to the motor cortex — no skull removal, no brain tissue incision, but a lower-resolution signal in exchange for a dramatically less invasive procedure.

Synchron trades signal fidelity for procedural simplicity. Neuralink pursues richer neural data through more invasive surgery. Neither approach has yet proven itself the clear winner.

What's actually been demonstrated, not promised

Neuralink had 21 patients worldwide living with an implant as of February 2026, up from a single recipient, Noland Arbaugh, when the company's human trials began. Its users have demonstrated controlling computers, smartphones, and robotic limbs using thought alone. Synchron's COMMAND study, involving six patients, met its primary safety endpoint with no device-related serious adverse events over 12 months, and in August 2025 one ALS patient in the trial publicly demonstrated navigating an iPad and composing text by thought alone, using Synchron's integration with Apple's BCI accessibility protocol, without any hand, voice, or eye movement.

The road from "it works" to "it's a product"

Both companies are now chasing the harder milestone: an actual approved, scalable product rather than a small trial. Synchron raised a $200 million Series D in November 2025 specifically to fund a 2026 pivotal trial aimed at filing the first premarket approval application for a permanently implanted BCI — a materially higher regulatory bar than the feasibility studies run so far. Neuralink, meanwhile, is pushing toward what Musk has called "high-volume production," including a more streamlined, largely automated surgical procedure intended to roll out in 2026. Competitors in the space have been blunt about what "high-volume" realistically means in the near term: hundreds of implants a year progressing toward thousands, not the tens of thousands Neuralink's own long-term framing suggests.

What to actually watch

The honest gap in this story isn't whether brain-computer interfaces work — they demonstrably do, for a small number of patients, right now. It's whether either company can clear the much higher bar of FDA approval at scale, manufacture and implant devices reliably outside a handful of specialized surgical centers, and do it while keeping safety data clean across a much larger patient population than the dozens involved so far. Synchron's 2026 pivotal trial and Neuralink's push toward an automated surgical procedure are both real steps in that direction — neither is a finish line yet.