Use case · AI
Approve AI decisions.
An AI agent is about to send an email, execute a payment, or sign a contract on your behalf. The question is not whether it can. The question is whether you can prove — years later — that you authorized this specific action under a specific policy.
The change: Every AI action carries a signed authority chain (governed AI authorization).
How it works
Four steps · every step signs a receipt.
Register an AI agent with scoped authority — what it can do, under what policy, for how long
Every time the agent acts, it presents its scope + the policy it's operating under · the runtime checks the chain before executing
Each executed action emits a signed receipt naming the agent, the authority, the policy, the evidence, and the preconditions
Later, any action is replayable from the receipt — you can reconstruct exactly why the agent was allowed to do this
Evidence
The proof this actually works.
- AI Agent Authorization is the production runtime for this pattern · every agent action already emits the 4-hash chain
- In-Vault replay lets you walk the chain step by step (see /replay/) — you can prove the decision was valid
- Object-Locked audit trail means the record cannot be deleted, even by an admin, for the retention period
- Edge case: An agent that goes rogue: the authority scope is bounded · it cannot exceed its grant even if compromised
- Edge case: Retroactive audit: policy at time-of-action is part of the receipt · you don't need to reconstruct 'what did the policy say six months ago'
Related capabilities
If you're comparing alternatives
Trust status
Runs today. Honest about what's next.
We publish what's shipped and what's coming — the same list on every page.
Runs today
- Vault workflows running in production
- Cryptographic receipts emitted per action
- Immutable audit trail · Object-Locked
- Auth1 · Wallet · Document · Audit APIs
- Post-quantum primitives (Dilithium + Kyber)
- Benchmark v12 · 2,293,766 auth/sec
Coming next
- Public /verify/
/ receipt viewer - Public /replay/
/ decision viewer - docs.h33.you developer documentation
- SOC 2 Type II Certified
- Payment Requests + Rewards adapters