Researcher Nathan Lambert has left the Allen Institute for AI after contributing to the Olmo model series. His departure marks the end of a tenure focused on open-source model development and research. This move shifts a known voice in the AI alignment and evaluation community away from the institute's direct internal operations.