Meta Backed Cortexa Wins FDA Nod For 100 WPM Non Invasive BCI, Big Tech Races To Embed Brain Input In Operating Systems
Cortexa Neural Interfaces, a privately held neurotech startup backed by Meta Platforms, has received De Novo clearance from the U.S. Food and Drug Administration for its high density non invasive brain computer interface to be used in communication for patients with ALS and other motor impairments.
In a pivotal 12 month study across 220 participants at major U.S. academic medical centers, the Cortexa Headset 1.0 enabled stable text entry at 70 to 100 words per minute and reliable cursor level control, using an array of 256 dry EEG electrodes embedded in a lightweight headset. The system relies on a set of transformer based decoding models trained on hundreds of thousands of hours of paired neural and behavioral data collected under a long running research collaboration with Meta's Reality Labs.
The device is paired with a palm sized signal processing unit that houses a custom low power ASIC designed by Cortexa and fabricated on a 12 nanometer process at Taiwan Semiconductor Manufacturing Company. The unit performs real time filtering and feature extraction before sending encrypted neural features over Bluetooth to a companion tablet or PC.
Alongside the FDA decision, Alphabet, Apple, and Microsoft each announced plans to ship native BCI input frameworks in their operating systems within the next 24 months. Alphabet said it will add a "NeuroInput" API layer to Android and ChromeOS, Apple previewed "NeuroKit" for iOS and visionOS, and Microsoft outlined a "Cognitive Input Framework" for Windows and Office 365.
Under these frameworks, BCI devices like Cortexa's headset will be treated as first class input peripherals on par with keyboards, mice, and touchscreens. Device makers will be able to register neural input channels with the OS, and applications will be able to request high level intents such as text, pointer movement, or menu selection without handling raw neural data.
Cortexa has signed long term supply agreements with key component vendors, including Infineon for mixed signal front end chips, Laird for electrode materials, and Flex for final device assembly. The company has also entered into distribution deals with Philips and Medtronic to target large hospital systems, and with Best Buy Health and Amazon to sell a "Cortexa Home" bundle for patients and caregivers.
Reimbursement will be critical for adoption. The Centers for Medicare and Medicaid Services is reviewing a proposal for new CPT codes that would cover BCI assisted communication therapy. Private payers, including UnitedHealth Group and Elevance, are watching the ALS data closely and have floated performance based agreements that tie reimbursement to measurable improvements in patient communication scores and reductions in caregiver burden.
Meta holds a significant minority stake in Cortexa and has a multi year IP and data sharing agreement that grants it preferential access to anonymized aggregate neural data and to certain decoding model improvements. In return, Meta has committed GPU resources and engineering support to help Cortexa scale training and inference infrastructure.
For hardware makers and software developers, the emergence of a cleared non invasive BCI device plus OS level support opens a new category of accessibility and productivity tools. For regulators and privacy advocates, it raises questions about how neural data should be governed, and who will set the standards for safe integration of brain signals into consumer devices.