Imagine being able to silently speak without moving a muscle. A bold Stanford study, published August 14, 2025, in Cell, just made that possible.

Researchers implanted electrodes in patients’ motor cortices and used AI to decode inner speech, allowing thoughts to be translated into words with up to 74% accuracy. This kind of technology could one day restore natural communication for people with conditions like ALS or paralysis, giving them a voice again.

Even more intriguing, the system sometimes picked up unprompted thoughts, which led the team to design a mental “password” safeguard. For example, thinking a phrase like “chitty chitty bang bang” could block private thoughts from being decoded.

This breakthrough not only opens a new chapter in brain-computer interfaces (BCIs) but also points toward a future where technology could restore voices once thought lost.