Max HendersonBoard member (founder of Ergo Impact)
Research institute · Research orgs
AVERI (the AI Verification and Evaluation Research Institute) is a 501(c)(3) US-based non-profit focused on enabling rigorous, independent auditing of “frontier” AI developers’ safety and security claims. AVERI’s approach is to combine (1) standards development, (2) pilot audits and audit-like evaluations intended to resolve open technical questions about audit access and confidentiality, and (3) technical/policy research translated into open tools, audit standards, and policy analysis/advocacy.
Publications are not yet indexed for this organization. Read them on averi.org ↗.
AVERI describes a pilot (July–August 2026) with Google DeepMind, OpenMined, and MLCommons, framed as a double-blind evaluation where prompts and model weights were protected via a secure enclave; the report discusses confidentiality constraints and proposed mechanisms (e.g., secure computation/TEEs).
Odd Lots Podcast appearance and commentaryAVERI lists an “Odd Lots Podcast” entry dated August 17, 2026.
Fortune Exclusive: Former OpenAI policy chief creates nonprofit institute, calls for independent safety audits of frontier AI modelsFortune coverage of AVERI’s launch and mission to enable independent safety/security audits for frontier AI models.