Researchers hack AI-enabled robots to cause ‘real world’ harm

Penn Engineering researchers said they created an algorithm that bypassed normal safety protocols stopping AI-powered robots from performing harmful actions.