Key points from article :
Researchers at UCLA have built a new brain-computer interface that doesn’t require surgery. The system uses an EEG cap to record brain signals and an AI program to read a person’s intent. The AI then helps steer a robotic arm or move a computer cursor with greater accuracy and speed.
The team tested the system on four participants, including one paralyzed volunteer. With AI assistance, all finished tasks much faster, especially the paralyzed participant, who could finally complete a block-moving task. Without the system, he could not perform it at all.
This design differs from older brain implants that needed risky neurosurgery. Wearable versions existed before, but they often lacked enough reliability for daily use. Combining EEG with AI solves this problem by acting as a “co-pilot,” guiding movements in real time.
Lead researcher Jonathan Kao, from UCLA’s Samueli School of Engineering, says the goal is independence. He explains the system could help people with paralysis or ALS regain control of everyday activities. His doctoral student, Johannes Lee, added that future versions may gain more precision and adaptability.
The study was carried out at UCLA with four test subjects. Results showed major improvements in speed and accuracy, highlighting the potential for safer, noninvasive support for those with mobility challenges.
The research was published in the journal Nature Machine Intelligence. It marks an important step toward practical brain-computer systems that patients can use outside of labs.
