House of ZK and Boundless, in partnership with Google Cloud, today announced Verifying Intelligence – a focused program on verifiable AI: proof-carrying inference, accountable data pipelines, audited agents, and privacy-preserving computation. The event, to be held on September 29th during TOKEN2049 week in Singapore, will convene prominent builders and policy leaders to define minimal verification standards for high-impact AI systems and to examine their societal externalities. Sponsors include Veridise, Ontology, 0G Labs, Laz AI, Gorka AI, and Altlayer.

Verifying Intelligence addresses a simple requirement: powerful systems must be provable. If model behavior, data lineage, and policy constraints cannot be demonstrated with evidence, the system should not be deployed. The program pairs concrete technical work – proof systems, attestations, and verification tooling – with discussion on real-world failure modes: misaligned optimization, synthetic media risk, power concentration, and the erosion of user agency.

“Verifiable AI should be viewed as an essential safety baseline,” said Ben Wynn, Chief Media Officer at House of ZK. “Our goal is to ultimately align builders on practical standards that any team can implement and any user can verify.”