Atlas / Digital minds / Paper

Studying AI Welfare Empirically

WelfareSubject & Substrate

Robert Long, Jeff Sebo, Patrick Butlin, Dillon Plunkett, Rosie Campbell, Charles Beasley, Bradford Saad, Toni Sims · 2026 · NYU CMEP and Eleos working paper

Read the source

A framework for the field: every welfare study should specify its question, the entity assessed (model, instance or persona) and the evidence type, and the research should be probabilistic, pluralistic, ethical, transparent and independent of AI companies.

Questions it bears on

Help improve this map

Know a paper that belongs here, or think something needs correcting?

Suggest a correction ↗Contributions are reviewed before they appear.