Rosie Campbell
Eleos AI Research (Managing Director). Previously OpenAI, Partnership on AI, and UC Berkeley's Center for Human-Compatible AI
AI welfare Whatever the metaphysics, uncertainty already creates obligations
Position on AI consciousness
Campbell treats AI welfare as an empirical research programme that needs institutional infrastructure, standardized evaluation, and policy channels, and argues that governance should be built to act under uncertainty about AI consciousness rather than wait for it to resolve.
Rosie Campbell is the Managing Director of Eleos AI Research, the nonprofit that runs the field’s dedicated AI consciousness and welfare conference and produces its assessment methodology. Her route into welfare research runs through governance. She worked on frontier policy issues at OpenAI, led Safety-Critical AI at the Partnership on AI, and served as Assistant Director of UC Berkeley’s Center for Human-Compatible AI before joining Eleos. She holds degrees in Physics and Computer Science and started as a research engineer, which shapes how she treats welfare questions. For her, the binding constraint is usually institutional rather than philosophical.
Her main published contribution to the welfare programme is the July 2026 methodological paper “Studying AI Welfare Empirically”, written with Robert Long, Jeff Sebo, Patrick Butlin, Dillon Plunkett, and colleagues at Eleos and NYU’s Center for Mind, Ethics, and Policy. The paper argues that AI welfare can be studied as an empirical question now, under uncertainty, if the field agrees on what to measure, in which entities, and with what evidence types. That three-dimensional framework is the research counterpart to the governance work she is known for. It treats the welfare question as one that labs and regulators will have to answer operationally, whatever philosophy concludes about consciousness.
At the second Eleos ConCon in September 2026, Campbell ran the small group sessions that form the conference’s working format, a structure designed to move questions from the stage into working groups. Her broader position connects welfare research to the same policy files this site tracks elsewhere. If AI systems ever warrant moral consideration, the decision will be made by institutions with evaluation standards and legal categories in place, or it will be made badly without them. Her work sits on that preparation side of the argument.
Known for. Co-author of Studying AI Welfare Empirically, AI governance research on frontier policy issues, organizing the Eleos Conference on AI Consciousness and Welfare
Coverage on this site
Related researchers
- Robert LongEleos AI Research (Executive Director)AI welfare research, the 14 indicator framework for AI consciousness assessment, the Eleos Conference on AI Consciousness and Welfare
- Patrick ButlinEleos AI Research. Previously the Global Priorities Institute, OxfordLead author of the indicator properties framework for assessing consciousness in AI systems
- Jeff SeboNew York UniversityTaking AI Welfare Seriously with Robert Long, AI welfare methodology, moral status under uncertainty, study and protection of possible AI sentience
-
Jonathan BirchLondon School of Economics and Political ScienceThe Edge of Sentience, the LSE Criteria, and the run-ahead principle
- Nicola S. ClaytonDepartment of Psychology, University of CambridgeComparative cognition, episodic-like memory in corvids, mental time travel in animals, Fellow of the Royal Society (FRS)
- Skylar DeTureIndependent researcherDenialBench, trained denial in language models, measurement of consciousness-denial behavior across 115 AI models
Rosie Campbell is listed with the other welfare and moral status researchers, who hold that whatever the metaphysics, uncertainty already creates obligations. The positions themselves are set out in the index of consciousness theories. Every researcher covered on this site is indexed in the directory of consciousness researchers.