ChoiceProof ProtocolA verifiable history
for AI that acts.
What was an AI agent authorized to do, and what did it actually execute? ChoiceProof Protocol links those facts in enterprise-attested receipts that can be checked later.
Make consequential
execution verifiable.
As AI moves into real operations, outcomes alone are not enough. We are building a shared way to connect authority, execution, and enterprise confirmation across organizational boundaries.
From action to verification.
Define the mandate
An enterprise sets the agent's identity, permitted actions, budget, provider scope, and expiration.
Record the execution
An authenticated connector or Agent API event is captured in a canonical execution receipt.
Attest at the enterprise
An enterprise signature is bound to the record, so ChoiceProof is not its only source of truth.
Verify the record
Inspect receipt hashes, enterprise attestation, optional blockchain anchoring, and audit export contents.
Published specification
The ChoiceProof Protocol v2.4.0 specification is finalized and published. It defines common rules for execution evidence, signatures, causal linkage, ordering, and independent verification.
The receipt hash can optionally be anchored to a blockchain for timestamp and tamper detection.
Read the developer specification