Research Scientist, Harmful Manipulation, DeepMind

DeepMind is seeking a Research Scientist to join its Responsible Development and Innovation (ReDI) team in New York, focusing on the Harmful Manipulation risk domain. The role involves researching emerging risks associated with AI-driven harmful manipulation, conducting technical investigations, and assessing how advanced AI models influence users. Responsibilities include designing and implementing evaluations to identify related capabilities and risks, as well as collaborating with cross-functional teams to perform comprehensive risk assessments. The position also entails publishing findings on novel methodologies and engaging with external stakeholders to inform governance decisions. Candidates should possess relevant skills and experience in AI research, risk assessment, and evaluation methodologies. While specific qualifications are not detailed, a strong background in these areas is implied. The job does not specify a working arrangement, leaving it open to interpretation. This role is ideal for individuals passionate about the intersection of AI safety, ethics, and practical application, who are eager to contribute to responsible AI development.