A recent study on brain-computer interfaces (BCIs) has found that surgically implanted devices can decode a person’s inner voice, even when they’re not actively speaking. The research, published in the journal Cell, suggests that BCIs could allow paralyzed users to produce synthesized speech more quickly and with less effort.
The study, led by Erin Kunz at Stanford University, used tiny electrode arrays to monitor brain activity in four people who were already using BCIs to communicate. The team found that imagined speech – thinking about a word or sentence without physically producing it – produces similar signals in the motor cortex as attempted speech.
With the help of artificial intelligence, the team was able to translate these fainter signals into words with an accuracy rate of 74%. This technology could revolutionize communication for paralyzed people, but raises concerns about privacy.
“The more we push this research forward, the more transparent our brains become,” says Nita Farahany, a professor of law and philosophy at Duke University. “Measures to protect people’s mental privacy are lagging behind technology that decodes signals in the brain.”
The study suggests that BCIs could be designed with safeguards to prevent unintentional sharing of private thoughts. One approach is to use a specific phrase or word to trigger decoding, as seen in virtual assistants like Alexa and Siri.
However, Farahany warns that even with these safeguards, the boundary between public and private thought may be blurred. “We have to recognize that this new era of brain transparency really is an entirely new frontier for us,” she says.
Source: https://www.npr.org/sections/shots-health-news/2025/08/20/nx-s1-5506334/brain-computer-implant-speak-inner-speech-mind