Meta Says New AI System Can Turn Non-Invasive Brain Activity Into Text

Meta has unveiled Brain2Qwerty, a new system that uses AI to translate brain activity into text from non-invasive brain recordings. According to the company, the model can turn recorded neural signals into sentences without surgery.

The development matters because it points to continued progress in AI-powered neural decoding, a field focused on interpreting brain signals through machine learning. Meta says its system improves accuracy compared with previous approaches, which could make brain-to-text technology more practical in research and future applications.

The company described the work as part of its broader efforts around AI and brain-signal interpretation. Unlike systems that require implanted devices, Brain2Qwerty relies on non-invasive recordings, which may lower barriers to testing and experimentation.

Meta’s announcement adds to growing interest in how AI can be used to interpret complex biological data. While the technology is still in the research phase, the company says the latest results show meaningful progress in translating neural activity into readable language.