← Back to Spotlight
Spotlight

Paralyzed Man Uses Brain-Computer Interface at Home to Speak, Work, and Browse the Web for Nearly Two Years

From Pepkio Team · 16 June 2026 · 3 min read

A 45-year-old man with severe paralysis and severe speech impairment from amyotrophic lateral sclerosis (ALS) has used an implanted brain–computer interface (BCI) in his own home to communicate, control a computer, and maintain full-time employment over a 19-month period. The study, published in Nature Medicine, shows that the system operated independently of research staff for more than 3,800 hours, allowing the participant to converse with family and colleagues, send emails, browse the internet, and join video calls.

The work was led by co-senior authors David M. Brandman at the University of California, Davis, and Sergey D. Stavisky, with first author Nicholas S. Card. Four microelectrode arrays were placed in the participant’s speech motor cortex as part of the BrainGate2 clinical trial.

During nearly two years of near-daily use at home, the participant generated over 183,000 sentences and nearly two million words at an average rate of 56 words per minute. He rated 92% of the decoded sentences as at least mostly correct. In formal testing where he was asked to repeat words shown on a screen, the speech decoder achieved more than 99% word accuracy from a vocabulary of 125,000 words. The same implant also decoded imagined hand movements to control a computer cursor, which he used to navigate his personal computer as mouse input alongside the speech-driven keyboard.

Critically, the system operated with minimal daily setup by the participant’s care partners and no researcher presence. After an initial training period, the BCI could operate continuously for up to 19 hours without additional assistance. The decoders adapted over time: an upgraded transformer-based speech model reduced the need for daily recalibration, and a recurrent neural network for cursor control shortened the time needed to gain cursor control to under a minute.

The neural signals themselves remained stable across the entire period. More than 90% of electrodes on each array continued to detect neural activity, and speech-related modulation patterns were highly similar even 18 months apart.

However, the study involved a single person, and the current system requires wires passing through the skin and a bulky cart, which limits portability. Speech accuracy during free conversation, while high, did not always match the near-perfect scores seen in structured tasks. Larger studies and fully implantable, wireless systems are the next steps.

“These results demonstrate that intracortical BCIs have the potential to support independent use in the home, marking a critical step toward practical assistive technology for people with severe motor impairment,” the authors write.

Reference:
Card, N.S., Singer-Clark, T., Peracha, H. et al. Long-term independent use of an intracortical brain–computer interface for speech and cursor control. Nat Med (2026). https://doi.org/10.1038/s41591-026-04414-6