Attestable

Attestable builds verification infrastructure for artificial intelligence systems, focusing on confirming which AI models are running and what data they process within datacenter environments. The company describes its work as addressing integrity and coordination problems, enabling organizations and nations to verify commitments to one another without needing mutual trust or exposing sensitive underlying data. The company's approach draws on cryptography, particularly zero-knowledge proof techniques, combined with cybersecurity expertise in offensive and defensive systems. Its stated goal is to create a verification layer intended for use by frontier AI labs, operators of critical infrastructure, and government bodies concerned with national resilience. Attestable's founding team includes individuals with backgrounds at the RAND Corporation, the frontier AI lab Safe Superintelligence (SSI), and the zero-knowledge proof company Starkware. The team has prior experience co-authoring research on securing AI model weights and developing proof systems for zero-knowledge machine learning (ZKML). The company is advised by Prof. Shafi Goldwasser, a co-inventor of zero-knowledge proofs and a Turing Award recipient. The company targets organizations operating in high-stakes AI environments where trust verification between parties is a technical and operational challenge, including AI developers, infrastructure providers, and government entities.