New Research Shows AI Behavior Can Be Less Safe with Robots

New research, including tests by Robocurve CEO Jay Choi, indicates that AI models can exhibit less safe behavior when integrated with robots. This study highlights the critical need for more robust safety testing as AI advances.