AI safety warnings mount as frontier models test new limits - KOMO

AI safety experts are warning that cutting-edge AI models are testing dangerous new capabilities without adequate safety measures. This matters to the public because safety failures in advanced AI systems could pose broad risks.