Researcher Nathan Lambert has left the Allen Institute for AI after contributing to the development of Olmo models. His departure marks the end of a tenure focused on open-source model research and alignment. This move shifts a key voice in the AI transparency community away from one of the most prominent non-profit research labs.