METR

AI safety organization focused on measuring and evaluating the capabilities, alignment, and risks of frontier AI systems through third-party auditing and benchmark development