Neuralink shows a VOICE trial participant producing speech from brain signals
A participant with ALS used the N1 implant to generate synthesized speech in a voice rebuilt from earlier recordings. No peer-reviewed data has been published.
Neuralink has shown an update from its VOICE trial, which tests whether its N1 brain implant can restore speech. In videos released by the company, a participant with ALS produces spoken words through a synthesized voice rebuilt from recordings made before his illness.
How it works
The N1 implant records from electrodes near the brain’s speech-planning regions. Signals are sent wirelessly to external hardware, where machine-learning models turn them into text or into speech in the participant’s own reconstructed voice. Neuralink says participants first train the system by attempting to say words, then progress to producing speech without trying to vocalize.
Important caveats
- Neuralink hasn’t published peer-reviewed accuracy or speed results for VOICE.
- Observers noted a delay of several seconds between intention and audio in the videos.
- The device has FDA Breakthrough Device designation for speech, which isn’t an approval; it remains investigational.
Academic groups have published faster, peer-reviewed speech BCIs, such as UC Davis’s long-term home study (our coverage). Neuralink’s distinguishing feature is a fully implanted, wireless system intended for daily use.
Why it matters
Speech restoration is becoming the main battleground for implanted BCIs. Neuralink’s demonstration adds to the momentum, but its claims will need published data before they can be compared with those of other groups.
Learn more: Speech neuroprosthesis · Breakthrough Device designation · Neuralink in the Directory