Work spans current risks such as misuse, bias, privacy, and security as well as risks from more capable systems. It uses technical safeguards, evaluation, operational controls, incident response, and policy.
AI safety is the study and practice of reducing harms from the design, deployment, and use of AI systems.
Work spans current risks such as misuse, bias, privacy, and security as well as risks from more capable systems. It uses technical safeguards, evaluation, operational controls, incident response, and policy.