Self-learning AI enhances behavioral analytics for detecting advanced threats and insider risks, but demands careful attention to bias, transparency, and ethical use.

Rose Yu has drawn on the principles of fluid dynamics to improve deep learning systems that predict traffic, model the climate, and stabilize drones during flight.

Scientists at Massachusetts Institute of Technology have devised a way for large language models to keep learning on the fly—a step toward building AI that continually improves itself.