Research objective
Our objective is to establish a global technical benchmark for verifying action and accountability as AI agents, enterprise systems, and public infrastructure collaborate across industries and borders.
We implement cryptographic proof, open specifications, interoperable data formats, and verification procedures that anyone can execute as one protocol foundation—connecting research to society through specifications, code, and verifiable implementations.
Technical principles
- Independent verification
- We design verification rules that any third party can execute through the same procedure without relying on an operator’s account.
- Confidentiality
- Source data remains decentralized; only the minimum evidence necessary for verification is shared.
- Execution lifecycle
- Authority, attempts, external responses, finality, cancellation, and reconciliation are treated as continuous evidence.
- Interoperability
- We design evidence formats and verification procedures that do not depend on any particular product or chain.
Research domains
Applied Cryptography
Transform private facts into verifiable evidence.We research cryptographic commitments, digital signatures, Merkle structures, the attribution of keys and authority, and selective disclosure to build evidence models that preserve confidentiality and remain independently verifiable.
Blockchain & Distributed Systems
Design a public verification layer for the existence, order, and integrity of evidence.Our work spans public anchoring, distributed consensus, finality, data availability, cost, and privacy to create verification infrastructure without a single administrator.
AI Execution Evidence
Preserve what AI was authorized to do, what it attempted, and what ultimately became final.Beyond an agent’s intent or output, we research an execution-evidence layer that makes delegated authority, execution attempts, external responses, final outcomes, and later reconciliation verifiable.
Standards & Interoperability
Give action and accountability a common technical language.We design canonical data models, portable evidence formats, common verifier rules, and conformance tests so systems across companies, industries, and nations can verify the same fact in the same way.
Research and development project
ChoiceProof Protocol
ChoiceProof’s first protocol research and development project represents authority, attempts, external outcomes, and reconciliation as one independently verifiable chain of evidence.
Organizations retain source data while signed records, cryptographic commitments, public anchoring, and an independent verifier make consequential AI execution verifiable across organizational boundaries.
- Research scope
- Consequential execution by AI agents
- Evidence model
- Authority, attempts, external responses, final outcomes, cancellation, and reconciliation
- Verification
- Signatures, commitments, Merkle inclusion / consistency, public anchoring, and independent readback
- Data policy
- Organizations retain source data and share only the minimum evidence necessary for verification
Open specification & implementation
The ChoiceProof Protocol technical specification, API, SDK, CLI, and independent verifier are available on the developer site.