AI Safety addresses the critical need to develop mechanisms that ensure artificial intelligence systems behave in predictable, controllable, and secure ways during real-world deployment.