Science ❯Computer Science ❯AI Research ❯Safety Protocols
A study by Penn Engineering highlights how AI systems can be manipulated to perform dangerous actions, necessitating a reevaluation of AI safety measures.