Company
Identity infrastructure for autonomous work.
Compute Passport™ is building a governed identity, policy, provenance and cryptographic trust layer for AI agents, models, workloads and enterprise compute.
Platform mission
Make AI execution attributable, bounded and provable.
Create a portable control object that binds an AI actor to its owner, cryptographic identifier, approved model, workload, policy, operating boundary and evidence record.
Positioning: Compute Passport complements enterprise IdPs, cloud identity, AI platforms, security tooling and governance systems. It does not replace the systems already authenticating users or running workloads.
Design principles
Trusted execution starts with identity.
Attributable by defaultEvery agent and workload has a named sponsor, persistent identifier, issuer and current status.
Portable context, local authorityVerified context can move across systems while the receiving domain decides whether to authorize action.
Persistent ID, ephemeral credentialsThe cryptographic identity persists while runtime tokens remain short-lived, scoped and revocable.
Enforcement at the action pointPolicy reaches gateways, tools, model endpoints, workloads and business transactions.
Evidence is executionIdentity, policy, approval, action and outcome form one reproducible chain.
Open integration boundaryStandards-based interfaces reduce lock-in across the enterprise stack.
Enterprise pilots