Internet-Draft Merkle Model Activation Binding November 2026
Durga Expires 6 May 2027 [Page]
Workgroup:
SCITT / RATS Working Group
Internet-Draft:
draft-durga-scitt-model-activation-binding-00
Published:
Intended Status:
Standards Track
Expires:
Author:
R. K. Durga
Microsoft Corporation

Merkle-Anchored Model Weight Release and RATS-Attested Runtime Activation Binding for Edge and Cloud AI Fleets

Abstract

Supply Chain Integrity, Transparency, and Trust (SCITT) provides transparent artifact statements, and Remote Attestation Procedures (RATS, RFC 9334) provide hardware/runtime attestation evidence. However, no existing internet standard defines an atomic cryptographic binding between a chunked AI model weight release manifest, a hardware-monotonic anti-rollback epoch counter, and a live RATS attestation result at the moment model weights are mapped into accelerator memory. This document specifies the Model Activation Binding Token (MABT) protocol, preventing load-time weight substitution, silent quantization downgrade, and rollback attacks across cloud and edge AI inference fleets.

Status of This Memo

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.

Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."

This Internet-Draft will expire on 6 May 2027.

▲

Table of Contents

1. Introduction and Problem Statement

AI model weights (GGUF, Safetensors, ONNX) are multi-gigabyte artifacts loaded dynamically by agent runtimes. Standard container image signing verifies the container binary at boot, but does not cryptographically bind the active model tensor chunks, LoRA adapter layers, and quantization profile in GPU/NPU memory to a SCITT transparency receipt and a RATS hardware nonce during inference execution.

This specification complements the foundational execution-trace verification architecture defined in draft-durga-agenttrust-trace-verification-00 by introducing deterministic, fail-closed wire semantics specifically addressing Merkle Model Activation Binding.

2. Conventions and Definitions

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 ([RFC2119] [RFC8174]) when, and only when, they appear in all capitals, as shown here.

3. Model Activation Binding Token (MABT) Structure

Prior to serving inference requests for a governed agent workflow, the inference runtime MUST generate and register a COSE_Sign1 or JCS-canonicalized Model Activation Binding Token (MABT):

{
  "mabt_version": "1.0",
  "workload_id": "spiffe://example.org/ns/prod/sa/agent-inference-node-04",
  "weight_merkle_root": "sha256:6b86b273ff34fce19d6b804eff5a3f5747ada4eaa22f1d49c01e52ddb7875b4b",
  "manifest_meta": {
    "chunk_size_bytes": 4194304,
    "total_byte_length": 16106127360,
    "quantization_profile": "FP16_STRICT"
  },
  "anti_rollback_security_epoch": 42,
  "rats_eat_attestation_hash": "sha256:d4735e3a265e16eee03f59718b9b5d03019c07d8b6c51f90da3a666eec13ab35",
  "scitt_transparency_receipt_ref": "sha256:4e07408562bedb8b60ce05c1decfe3ad16b72230967de01f640b7e4729b49fce",
  "activation_nonce": "9f86d081884c7d659a2feaa0c55ad015",
  "exp": 1793491200,
  "signature": {
    "alg": "Ed25519",
    "kid": "sha256:b600306cfa76723fdec395e53a9b3d9fdb78b1e2d7a23c32fcbcd2dc6d0c4092",
    "sig": "b3cCTQWySKBHRIbG2jPGHgJXcAsxYhAXLZ6sYdn_vhQrhViPT48ZRjLk-yG3N0A1Aj-XDJspWC6eT4jryzLYAA"
  }
}

4. Normative Fail-Closed Verification Rules

Conforming verifiers and relying-party gateways MUST enforce the following deterministic invariants before releasing any downstream side effect:

  1. CHUNKED_WEIGHT_MERKLE_VERIFICATION: The weight_merkle_root MUST be computed as a binary Merkle tree (RFC 9162 style domain separation: 0x00 for leaves, 0x01 for internal nodes) over ordered, fixed-size tensor chunks prior to DMA transfer into accelerator memory.
  2. ANTI_ROLLBACK_EPOCH_ENFORCEMENT: Conforming verifiers MUST reject any MABT whose anti_rollback_security_epoch is strictly less than the minimum security epoch persisted in hardware NVRAM or the fleet policy ledger.
  3. RATS_NONCE_FRESHNESS: The rats_eat_attestation_hash MUST bind the runtime's hardware measurement, weight_merkle_root, and activation_nonce so replayed attestations from prior boot cycles fail closed.

5. Strict Parser Determinism and Wire-Level Input Hardening

Cross-language interoperability testing across Rust and Python attestation verifiers demonstrates that standard RFC 8785 JCS canonicalization is necessary but insufficient if the outer JSON/CBOR parser silently coerces malformed inputs prior to canonicalization. Conforming MABT verifiers MUST enforce five pre-canonicalization parser invariants:

  1. DUPLICATE_KEY_REJECTION: The JSON parser MUST reject any MABT object or nested manifest_meta map containing duplicate member names at the lexer/object-pairs hook prior to dictionary construction, preventing last-key-wins shadow payload attacks.
  2. IEEE754_FINITE_SAFE_INTEGER_BOUNDS: All integer fields (chunk_size_bytes, total_byte_length, anti_rollback_security_epoch, exp) MUST be exact JSON integers within the IEEE-754 safe integer range [-(2^53 - 1), +(2^53 - 1)]. Boolean literals (true/false), fractional numbers, and non-finite constants (NaN, Infinity, -Infinity, or exponent-overflow tokens such as 1e1000) MUST be rejected immediately.
  3. CANONICAL_LOWERCASE_HEX_DIGESTS: All sha256:-prefixed digest fields (weight_merkle_root, rats_eat_attestation_hash, scitt_transparency_receipt_ref, kid) MUST consist of the literal lowercase prefix sha256: followed by exactly 64 lowercase ASCII hexadecimal characters ([0-9a-f]{64}).
  4. MINIMAL_BASE64URL_AND_PUBLIC_JWK_HYGIENE: The sig field MUST be unpadded RFC 4648 Section 5 Base64url with zero non-zero trailing pad bits. Any embedded or referenced JWK key material MUST reject private key parameters (d, p, q, dp, dq, qi, oth, k) and non-minimal Base64urlUInt RSA exponents.
  5. EVIDENCE_TO_RULE_BYTE_BINDING: When a MABT is cited as an evidence item inside an external Verdict Record ([I-D.wright-verdict-record]), the verifier rule evaluation MUST bind directly to the digest-verified MABT envelope bytes rather than unverified summary copies in the outer claim body.

6. Implementation Status and Empirical Benchmark Evaluation

The normative mechanisms and strict parser hardening rules specified in this document have been implemented and evaluated across open-source reference verifiers and production attestation pipelines:

7. Security Considerations

All pre-image objects MUST be canonicalized using RFC 8785 JSON Canonicalization Scheme (JCS) prior to signature generation or verification under RFC 8032 Ed25519. Any clock skew exceeding 60 seconds, digest mismatch, or missing cryptographic binding MUST cause the verifier to fail closed and emit an auditable security rejection receipt. For the normative test vector in Section 3, the Ed25519 test key seed is 2a2a...2a (32 bytes of 0x2a), yielding raw 32-byte public key 197f6b23e16c8532c6abc838facd5ea789be0c76b2920334039bfa8b3d368d61 (kid = sha256:b600306cfa76723fdec395e53a9b3d9fdb78b1e2d7a23c32fcbcd2dc6d0c4092), and the full RFC 8785 JCS-canonicalized envelope (822 bytes) hashes to SHA-256 f12feb5bf217115ee22b4e82b9da13de2497000fddd248ebeaec6d4969b405c1. This specification cross-references the companion Internet-Drafts ([I-D.durga-agenttrust-trace-verification], [I-D.durga-wimse-ctx-isolation], [I-D.durga-pquip-pqc-agentic-delegation], [I-D.durga-agent-tx-compensation-lease]), external Verdict Record evidence binding ([I-D.wright-verdict-record]), Remote Attestation ([RFC9334]), Merkle Tree domain separation ([RFC9162]), and the 100-scenario OWASP ASI01-ASI10 benchmark ([OWASP-ASI-BENCHMARK-DOI]).

8. IANA Considerations

This document requests IANA registration of the media type application/mabt-activation+json and the registry IETF Model Activation Verification Profiles, alongside the following header fields:

9. Normative References

[RFC2119]
Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, , <https://www.rfc-editor.org/info/rfc2119>.
[RFC8174]
Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words", BCP 14, RFC 8174, , <https://www.rfc-editor.org/info/rfc8174>.
[RFC8785]
Rundgren, A., Jordan, B., and S. Erdtman, "JSON Canonicalization Scheme (JCS)", RFC 8785, , <https://www.rfc-editor.org/info/rfc8785>.
[RFC8032]
Josefsson, S. and I. Liusvaara, "Edwards-Curve Digital Signature Algorithm (EdDSA)", RFC 8032, , <https://www.rfc-editor.org/info/rfc8032>.

10. Informative References

[RFC9162]
Laurie, B., Messeri, E., and R. Stradling, "Certificate Transparency Version 2.0", RFC 9162, , <https://www.rfc-editor.org/info/rfc9162>.
[RFC9334]
Birkholz, H., Thaler, D., Richardson, M., Smith, N., and W. Pan, "Remote ATtestation procedureS (RATS) Architecture", RFC 9334, , <https://www.rfc-editor.org/info/rfc9334>.
[I-D.durga-agenttrust-trace-verification]
Durga, R. K., "Cryptographic Execution-Trace Verification, JCS Canonicalization, and Fail-Closed Delegation Chains for Autonomous AI Agents", Work in Progress, Internet-Draft, draft-durga-agenttrust-trace-verification-00, , <https://datatracker.ietf.org/doc/html/draft-durga-agenttrust-trace-verification-00>.
[I-D.durga-wimse-ctx-isolation]
Durga, R. K., "Cryptographic Context Isolation and Instruction-Data Plane Separation for Autonomous Agent Tool Invocations", Work in Progress, Internet-Draft, draft-durga-wimse-ctx-isolation-00, , <https://datatracker.ietf.org/doc/html/draft-durga-wimse-ctx-isolation-00>.
[I-D.durga-pquip-pqc-agentic-delegation]
Durga, R. K., "Post-Quantum Hybrid Cryptographic Delegation Chains and Merkle-Aggregated Execution Receipts for Autonomous AI Agents", Work in Progress, Internet-Draft, draft-durga-pquip-pqc-agentic-delegation-00, , <https://datatracker.ietf.org/doc/html/draft-durga-pquip-pqc-agentic-delegation-00>.
[I-D.durga-agent-tx-compensation-lease]
Durga, R. K., "Pre-Release Transactional Compensation Leases and CAS Fencing Receipts for Autonomous Agents Over Systems of Record", Work in Progress, Internet-Draft, draft-durga-agent-tx-compensation-lease-00, , <https://datatracker.ietf.org/doc/html/draft-durga-agent-tx-compensation-lease-00>.
[I-D.wright-verdict-record]
Wright, D., "Verdict Records: Results on Claims that Cite Receipts as Evidence", Work in Progress, Internet-Draft, draft-wright-verdict-record-01, , <https://datatracker.ietf.org/doc/html/draft-wright-verdict-record-01>.
[OWASP-ASI-BENCHMARK-DOI]
Durga, R. K., "OWASP ASI01-ASI10 Enterprise Agentic AI Security Regression Benchmark Dataset (N=100 Scenarios)", DOI 10.57967/hf/11352, , <https://doi.org/10.57967/hf/11352>.
[XERJ-HUB-EVIDENCE]
Durga, R. K., "XERJ Common Rust Engine: Deterministic JCS TraceEvidenceReport, PinnedCitation, and Isotonic Confidence Verification", GitHub xerj-org/xerj#2857, , <https://github.com/xerj-org/xerj/pull/2857>.
[AGENTIC-GOV-BENCHMARKS]
Durga, R. K., "Enterprise Agentic Governance Benchmarks and Reference Verifier Suite v1.0.0", Version 1.0.0, , <https://github.com/drk-ram/enterprise-agentic-governance-benchmarks/releases/tag/v1.0.0>.
[ZENODO-TR-2026-01]
Durga, R. K., "Cryptographic Model Weight Release, Hardware Anti-Rollback Leases, and RATS-Attested Activation Binding for Edge AI Fleets", DOI 10.5281/zenodo.23735274, , <https://doi.org/10.5281/zenodo.23735274>.

Author's Address

Raghuvar Karthik Durga
Microsoft Corporation
United States of America