Meta said its non-invasive brain-to-text system reached 61% accuracy, underscoring both its promise for assistive communication and the privacy concerns tied to decoding neural signals.
Meta has unveiled AI technology designed to translate brain activity into text using a non-invasive approach, with the system reported to achieve 61% accuracy. The development could improve communication for people with severe motor impairments by converting brain signals into written language without implanted devices. The advance also raises privacy concerns, as more capable neural decoding systems could sharpen questions about how sensitive brain data is collected, used and protected.