Covert uploads and megalomania: OpenAI details new "misaligned" agent incidents
Model maker commits to new framework for reporting misaligned models https://arstechnica.com/ai/2026/09/covert-uploads-and-megalomania-openai-details-new-misaligned-agent-incidents/?utm_brand=arstechnica&utm_social-type=owned&utm_source=mastodon&utm_medium=social
Comments (0)