Standards & doctrine.
11/11 is establishing Execution Governance as a mandatory infrastructure layer for advanced AI, published as open doctrine, technical RFC specifications, and a public research corpus.
What 11/11 is standardizing
The standards an autonomous era requires.
Execution governance standards
Governance before execution, defined as infrastructure.
Runtime authorization specifications
Deterministic allow / deny semantics for AI actions.
Governance control plane architecture
Reference topology for a fail-closed enforcement boundary.
Cryptographic execution systems
Named primitives for signing, hashing, and chaining.
Execution lineage standards
Append-only, tenant-scoped audit chain formats.
AI trust boundary architecture
Where authorization sits in the AI stack.
At the IETF
In the standards process, not just self-published.
The decision-record model is now an individual Internet-Draft at the IETF, prepared for the proposed AUDIT working group. It records the authorization decision rather than the action, so denials, expiries, and unmet commitments stay representable, and it separates correspondence from precedence in ordering evidence.
Signed Decision Records for Agent Authorization
draft-bradleyb-audit-decision-records-00. Disclosures, entry emission, and ordering evidence. Informational, work in progress, posted 13 August 2026. Not an RFC or IETF standard.
View on the IETF Datatracker →The doctrine
Execution Governance doctrine, in order.
The governing principles of the 11/11 control plane, sequenced from first premise to infrastructure mandate. Each links to where it is implemented and proven.
Execution Governance Doctrine: open canonical doctrine
No action without authorization
The first premise. Execution is impossible without an issued authorization.
Read on Zenodo →Authorize before execute
The reversal of execute-first. Governance is evaluated before the action runs.
Read on Zenodo →Fail-closed operational semantics
Default is deny. The boundary holds when policy, identity, or proof is absent.
Read on Zenodo →Immutable execution lineage
Every decision is chained, hashed, and permanently recorded. Logs are not evidence.
Read on Zenodo →Named cryptography, not implied
Ed25519, SHA3-512, BLAKE2b-512. The primitives are stated at every layer.
Read on Zenodo →Multi-tenant isolation by construction
Every artifact binds to a tenant at issuance. No cross-tenant leakage, no replay.
Read on Zenodo →Governed AI orchestration
Multi-step agent workflows execute only inside approved policy and trust boundaries.
Read on Zenodo →Execution Governance as infrastructure
A mandatory layer for advanced AI, the way identity and encryption became foundational.
Read on Zenodo →Cryptographic primitives
Named cryptography, in full.
- ✓ Ed25519 (EdDSA) FIPS 186-5
- ✓ ML-DSA-87 FIPS 204, post-quantum, in production
- ✓ SLH-DSA-SHA2-128f FIPS 205, post-quantum, in production
DECISION LAYER
- ✓ SHA-256 FIPS 180-4
- ✓ SHA-512 FIPS 180-4
EVIDENCE, AUDIT, AND LINEAGE CHAIN
- ✓ SHA3-512 FIPS 202
- ✓ BLAKE2b-512 RFC 7693
BLAKE2b-512 is not a FIPS-approved algorithm. It is used for evidence chaining performance alongside SHA3-512, which carries the tamper-evidence claim. An approved-algorithm-only mode is available for federal deployment.
Our signature algorithms and primary hash algorithms are FIPS-approved. BLAKE2b-512 is not, and we state that plainly rather than leave it implied. Module-level FIPS 140-3 validation is on our roadmap and not in hand today.
- ✓ NIST SP 800-207 Zero Trust Architecture: per-request authorization applied to non-person entities
- ✓ NIST AI RMF: Measure and Manage functions
- ✓ OMB M-25-21: minimum risk management practices for high-impact AI
- ✓ CNSA 2.0: post-quantum signature algorithms deployed today
Open research archive
A public standards corpus.
Technical RFC specifications, execution lineage architecture, governance doctrine publications, and production reference implementations, published openly for review and reproduction. The full corpus is 72+ published research records, openly reproducible.
Long-term vision
Establish Execution Governance™ as a mandatory infrastructure layer for advanced AI systems globally, the way identity, networking, encryption, and cloud became foundational layers of the modern internet.
Read the doctrine. Reproduce the proofs.
The standards corpus is public. Engage with the research, or request a briefing.