Technology breakthrough as brain implant turns thoughts to speech in 3 seconds (2025)

A new brain chip has allowed a paralysed woman to turn her thoughts into a voice

In a groundbreaking scientific breakthrough, a woman who lost her ability to speak due to paralysis can now "speak" again – with the help of a revolutionary brain implant that translates thoughts into speech in less than three seconds.

The brain-reading technology, known as a brain-computer interface (BCI), uses a high-tech implant to read signals from the brain and instantly turn them into audible speech.

The implant picks up on each word as it's being thought and speaks it out loud through a voice personalised to sound like the patient's own voice before her injury.

The patient, referred to in the study as Ann, suffered a stroke in 2005 that left her unable to speak. After undergoing surgery to implant a device into her brain, she can now "speak" – and the device's artificial intelligence (AI) was even trained to mimic her real voice from old wedding videos.

“Our streaming approach brings the same rapid speech decoding capacity of devices like Alexa and Siri to neuroprostheses,” said Gopala Anumanchipalli from the University of California Berkeley, an author of the study.

“Using a similar type of algorithm, we found that we could decode neural data and, for the first time, enable near-synchronous voice streaming. The result is more naturalistic, fluent speech synthesis," Anumanchipalli added.

Previous versions of this technology could only generate sounds after an entire sentence was finished, making communication feel slow and clunky. Now, the system can quickly translate thoughts into speech at up to 90 words per minute – nearly double the speed of previous models which also had around an eight-second delay for a single sentence.

In the study, the researchers showed Ann 100 sentences from a set of more than 1,000 words and 50 phrases on a screen. As she silently mouthed the sentences, the device detected her neural signals every 80 milliseconds, generating between 47 and 90 words per minute of speech.

To test whether the AI was successfully "learning", the researchers showed Ann words that were not part of the training dataset.

“We wanted to see if we could generalize to the unseen words and really decode Ann’s patterns of speaking,” Anumanchipalli said. “We found that our model does this well, which shows that it is indeed learning the building blocks of sound or voice.”

While it's still not as fast as a natural conversation, which happens at around 160 words per minute, researchers believe the technology is only going to get faster and smoother.

“This proof-of-concept framework is quite a breakthrough,” said study co-lead author Cheol Jun Cho, a PhD student at Berkeley. “We are optimistic that we can now make advances at every level. On the engineering side, for example, we will continue to push the algorithm to see how we can generate speech better and faster.”

Neurosurgeon and co-principal investigator of the study, Edward Chang, described the technology as "exciting".

“This new technology has tremendous potential for improving quality of life for people living with severe paralysis affecting speech,” he said. “It is exciting that the latest AI advances are greatly accelerating BCIs for practical real-world use in the near future."

The study was published in the journal Nature Neuroscience.