Network Working Group D. Watts Internet-Draft ai-id-engine project Intended status: Experimental 16 November 2026 Expires: 20 May 2027 An Entity Attestation Token (EAT) Profile for Signed Pre-Silicon Hardware Security Evidence draft-watts-rats-presilicon-evidence-00 Abstract Hardware security designs are increasingly assessed before fabrication: by gate-level leakage simulation, fault-injection simulation, formal proofs and equivalence checking. The results of these assessments are routinely copied into reports, slide decks and certification packages with no binding to the design that was assessed, the stimulus that was used, the conditions under which the numbers hold, or the platform that produced them. This document defines an Entity Attestation Token (EAT) profile for carrying such pre-silicon assessment results as signed Evidence. Every claim is classified as identifying the assessed artifact, describing a measured behavior, or describing the measurement conditions. A provenance claim, which states whether the signing key is rooted in hardware or software, is mandatory, and a simulated Trusted Platform Module is always reported as a software root. The profile is explicitly not a certification format: it records what was measured and under which envelope, so that a Relying Party can decide what the numbers are evidence of. About This Document This note is to be removed before publishing as an RFC. A reference implementation (schema "aiid-evidence/v1", JSON encoding, TPM 2.0 or software-key signing) is in the ai-id-engine repository, files hwid/src/evidence.rs and docs/EVIDENCE.md. This draft describes that implementation and proposes how it maps onto EAT. It is experimental. Status of This Memo This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79. Watts Expires 20 May 2027 [Page 1] Internet-Draft Pre-Silicon Evidence Profile November 2026 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 20 May 2027. Copyright Notice Copyright (c) 2026 IETF Trust and the persons identified as the document authors. All rights reserved. This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/ license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 1.1. Requirements Language . . . . . . . . . . . . . . . . . . 4 1.2. Terms . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2. Evidence Model . . . . . . . . . . . . . . . . . . . . . . . 4 2.1. Required Fields . . . . . . . . . . . . . . . . . . . . . 5 2.2. Controls Are Part of the Result . . . . . . . . . . . . . 6 2.3. Measurement Envelope . . . . . . . . . . . . . . . . . . 6 3. Provenance . . . . . . . . . . . . . . . . . . . . . . . . . 7 4. Signing and Binding . . . . . . . . . . . . . . . . . . . . . 7 5. Mapping to EAT . . . . . . . . . . . . . . . . . . . . . . . 8 6. Appraisal Guidance . . . . . . . . . . . . . . . . . . . . . 8 7. Example . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 8. Security Considerations . . . . . . . . . . . . . . . . . . . 9 9. Privacy Considerations . . . . . . . . . . . . . . . . . . . 10 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 10 10.1. EAT Profile Identifier . . . . . . . . . . . . . . . . . 10 10.2. Claims . . . . . . . . . . . . . . . . . . . . . . . . . 11 11. References . . . . . . . . . . . . . . . . . . . . . . . . . 11 11.1. Normative References . . . . . . . . . . . . . . . . . . 11 11.2. Informative References . . . . . . . . . . . . . . . . . 11 Implementation Status . . . . . . . . . . . . . . . . . . . . . . 12 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . 12 Watts Expires 20 May 2027 [Page 2] Internet-Draft Pre-Silicon Evidence Profile November 2026 1. Introduction The RATS architecture [RFC9334] describes how an Attester produces Evidence about itself, how a Verifier appraises that Evidence, and how a Relying Party consumes the resulting Attestation Results. The Entity Attestation Token [RFC9711] provides a claims-based encoding for Evidence and allows profiles that constrain and extend it for a particular use. This document addresses a use that existing profiles do not cover: Evidence about a hardware design, produced before that design exists as silicon. Typical sources are: * Test Vector Leakage Assessment (TVLA) [TVLA] on simulated, gate- level power traces of a cryptographic block; * fault-injection simulation of error-detecting codes and state encodings; * formal proofs of safety and non-interference properties; * equivalence checks between register-transfer-level (RTL) source and a synthesized netlist; * reproducible-build digests of an FPGA bitstream used as a prototype. Such results are only meaningful relative to the exact netlist, cell library, power model, stimulus set and statistical parameters that produced them. A first-order |t| value of 3.15 says nothing on its own. Together with "4000 traces per group, 4000 samples, alpha 0.05, Bonferroni threshold 4.37, sky130 typical corner, zero-delay cell models, noiseless", it becomes a reproducible statement with a known minimum detectable effect. This profile binds the numbers to those conditions under a signature, and records the trust root of that signature honestly. This profile does not define how to perform any of these assessments, and it does not turn their results into a certification. A report that conforms to this profile can be evidence for a certification process; it is never itself a certificate. Watts Expires 20 May 2027 [Page 3] Internet-Draft Pre-Silicon Evidence Profile November 2026 1.1. Requirements Language 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. 1.2. Terms Assessed artifact: The design under assessment: RTL source, synthesized netlist, derived power model, cell models and cell library. Assessment host: The machine that ran the assessment and signs the report. In RATS terms it is the Attester. Its Target Environment is the assessment run, not the assessed artifact. Measurement envelope: The set of modelling assumptions under which the behavior claims hold, together with an explicit list of effects that were not modelled. Provenance root: "hardware" if the report signature is produced by a key held in a physical TPM; "software" otherwise. 2. Evidence Model A report is a body of named fields plus an attestation over the body. Each field carries exactly one class: identity The field identifies what was assessed or who signed: artifact digests, the assessment host's identifier, and provenance. behavior The field is a measured outcome: leakage statistics, a gate verdict, area, latency, fault-detection counts, proof status. neither The field describes the conditions of the measurement: statistical parameters, the stimulus set, the measurement envelope, tool versions. The classification is part of the signed body. A Verifier MUST reject a report in which: * any field lacks a class, or has a class other than these three; Watts Expires 20 May 2027 [Page 4] Internet-Draft Pre-Silicon Evidence Profile November 2026 * a required field (Section 2.1) is missing or carries the wrong class; * a field name appears more than once. The classes let a Relying Party apply policy mechanically. For example, it can accept behavior claims only when the identity claims match a netlist digest it has independently obtained, and only when the neither claims show a trace count at or above its own minimum. 2.1. Required Fields +======================+==========+===============================+ | Name | Class | Content | +======================+==========+===============================+ | provenance | identity | Signing root; see Section 3. | +----------------------+----------+-------------------------------+ | hwid | identity | Salted identifier of the | | | | assessment host. | +----------------------+----------+-------------------------------+ | model_digest | identity | Map from artifact path to | | | | SHA-256 [FIPS180-4] digest. | | | | It covers the RTL, netlists, | | | | power-model netlists and | | | | capacitance files, cell | | | | models and the cell library. | +----------------------+----------+-------------------------------+ | stimulus_set | neither | Digests of the stimulus | | | | generator, harness and | | | | configuration, plus seeds, | | | | modes and trace counts per | | | | group. | +----------------------+----------+-------------------------------+ | tvla_parameters | neither | Profile name, assessed | | | | design, traces per group, | | | | samples per trace (M), alpha, | | | | the first- and second-order | | | | critical values, and the | | | | number of second-order tests. | +----------------------+----------+-------------------------------+ | measurement_envelope | neither | See Section 2.3. | +----------------------+----------+-------------------------------+ | tvla_result | behavior | Null-control and positive- | | | | control peaks, first- and | | | | second-order peaks, minimum | | | | detectable effect, verdict | | | | and gate exit code. | +----------------------+----------+-------------------------------+ Watts Expires 20 May 2027 [Page 5] Internet-Draft Pre-Silicon Evidence Profile November 2026 Table 1: Fields every conforming report carries Additional fields MAY be present (for example area, latency, fault- injection counts or equivalence status). Each additional field MUST carry a class. 2.2. Controls Are Part of the Result A leakage statistic without controls cannot be interpreted. The null control (random versus random inputs on the assessed design) bounds the false-positive behavior of the harness. The positive control (fixed versus random on an unprotected reference) shows that the harness can see a leak. A conforming tvla_result MUST include both control results. A Verifier MUST NOT treat a report as evidence of absence of leakage if the null control exceeded its threshold or the positive control did not. 2.3. Measurement Envelope The measurement_envelope states what the behavior claims are evidence of. It MUST contain: * a kind string that says whether the traces were simulated or measured; for simulated traces it MUST contain the words "not silicon"; * the cell library and process corner; * the power model and the wire-load model; * the measurement chain parameters (for example low-pass time constant and additive noise); * samples per trace; * a list of effects not modelled (for example glitches, coupling, IR drop, electromagnetic emanation, process variation, parasitics after placement and routing); * tool names and versions. The list of effects not modelled is REQUIRED because it is the part of a pre-silicon report most often lost when results are quoted elsewhere. Watts Expires 20 May 2027 [Page 6] Internet-Draft Pre-Silicon Evidence Profile November 2026 3. Provenance The provenance field records the root of the key that signs the report. Its value contains: * attestation_kind: "tpm2-quote" or "software-fallback"; * source: how the platform information was obtained, for example TPM2_GetCapability for TPM_PT_MANUFACTURER and TPM_PT_VENDOR_STRING_1 [TPM2], or "none" when no TPM was used; * tpm_manufacturer, and a simulated_tpm flag; * root: "hardware" or "software". The following rules apply: * root MUST be "hardware" only when the signature is a TPM 2.0 quote from a physical TPM. * A software TPM (for example swtpm or the Microsoft simulator) MUST be reported as root "software", even though it speaks the TPM 2.0 protocol. * A plain software key MUST be reported as root "software". * The Attester obtains the provenance before signing and places it in the body, so the signature covers it. * A Verifier MUST recompute the expected provenance from the attestation it verified, and MUST reject the report if the two differ. In particular it rejects a body that claims "hardware" over a simulated TPM quote. 4. Signing and Binding The body is serialized canonically: compact JSON [RFC8259] with fields in the order written. The report nonce is: nonce = SHA-256( "AIID-EVIDENCE-V1" || canonical(body) ) The nonce is used as the freshness value of the platform attestation. * For a TPM, the quote's qualifying data is derived from the nonce and the host identifier, and the quote signs it with an Attestation Key whose binding to the Endorsement Key was checked by credential activation. Watts Expires 20 May 2027 [Page 7] Internet-Draft Pre-Silicon Evidence Profile November 2026 * For a software key, the key signs the nonce together with the host identifier and a timestamp. The report is the triple (body, body hash, attestation). A Verifier MUST: 1. validate the body against Section 2; 2. recompute the canonical body hash and compare it; 3. verify the attestation over the recomputed nonce; 4. check provenance agreement as described in Section 3. 5. Mapping to EAT This section describes how the model maps onto an EAT [RFC9711] profile, so that reports can travel in RATS conceptual messages. * The eat_profile claim MUST be set to the profile identifier in Section 10.1. * The eat_nonce claim carries the nonce defined in Section 4. * Each body field is carried as a claim in a presilicon-evidence claims-set. The claim value is a map with keys "class" and "value". * The provenance field maps onto the existing hardware and software model claims where those apply. It MUST also be carried verbatim, because EAT has no claim for "simulated TPM". * When the attestation is a TPM quote, the token is carried together with the quote in the evidence bundle, as for other TPM-based Evidence. A software-root report SHOULD be signed with COSE or JOSE using the software key, and still declares root "software". 6. Appraisal Guidance A Verifier producing Attestation Results from such Evidence SHOULD: * report the provenance root unchanged; * report the trace count and the minimum detectable effect beside any leakage verdict; * never translate a passing verdict at a given trace count into a statement about leakage at higher trace counts or on silicon. Watts Expires 20 May 2027 [Page 8] Internet-Draft Pre-Silicon Evidence Profile November 2026 A Relying Party SHOULD compare model_digest against artifacts it obtained independently, for example from a reproducible build whose digests are themselves published. 7. Example The following abbreviated body comes from the reference implementation. It describes a 2-share masked Montgomery multiplier at 4000 traces per group, which fails second-order TVLA as expected. Digests are truncated. {"schema":"aiid-evidence/v1","created":"2026-11-16T16:10:00Z", "fields":[ {"name":"provenance","class":"identity","value":{ "attestation_kind":"software-fallback", "source":"none: no TPM, software-fallback key file", "tpm_manufacturer":"","simulated_tpm":false,"root":"software"}}, {"name":"hwid","class":"identity","value":"9f2c...e1"}, {"name":"model_digest","class":"identity","value":{ "build/sky130_masked_v2.v":"4b1e...07", "build/tvla_masked_v2.caps":"c3a9...5d"}}, {"name":"stimulus_set","class":"neither","value":{ "files":{"config/tvla.toml":"77d0...2a"}, "description":{"fixed_vs_random":{"seed":17, "traces_per_group":4000}}}}, {"name":"tvla_parameters","class":"neither","value":{ "profile":"ci","traces_per_group":4000,"samples":4000, "alpha":0.05,"crit_first_order":4.37,"crit_second_order":4.46}}, {"name":"measurement_envelope","class":"neither","value":{ "kind":"simulated gate-level power traces; not silicon", "library":"sky130_fd_sc_hd__tt_025C_1v80", "not_modelled":["glitches","coupling and IR drop","EM"]}}, {"name":"tvla_result","class":"behavior","value":{ "null_control":{"peak":3.24,"crit":4.37,"passed":true}, "positive_control_plain":{"peak":82.96,"detected":true}, "first_order_peak":3.15,"second_order_peak":26.50, "mde_sigma":0.0977,"exit":2, "verdict":"second-order leak in 2-share design"}}]} 8. Security Considerations This profile improves the binding and honesty of pre-silicon results. It does not make them stronger than the assessment that produced them. Simulation is not silicon. Zero-delay gate-level simulation does not model glitches, coupling, IR drop or electromagnetic emanation, and second-order Watts Expires 20 May 2027 [Page 9] Internet-Draft Pre-Silicon Evidence Profile November 2026 statistics may be validated only synthetically. A Relying Party MUST NOT treat a conforming report as a side-channel certification. Trace count bounds what a pass means. A pass at N traces per group shows only that no leak larger than the stated minimum detectable effect was found at N. Provenance is host-reported. The platform information comes from the host's TPM software stack. It is covered by the Attestation Key signature, but a compromised host can misreport the TPM vendor unless the Verifier also validates the Endorsement Key certificate chain, which the reference implementation does not yet do. A software root is a file key. It proves possession of a key file, not the integrity of the host. Reports from continuous-integration systems will typically have a software root. Digests bind; they do not judge. model_digest ties the results to particular files, but the Relying Party must still establish that those files are the design it cares about. Replay. The nonce is derived from the body, so the same body always yields the same nonce. Verifiers that need freshness SHOULD also check the attestation timestamp or request a report bound to their own challenge. 9. Privacy Considerations The hwid field is a salted identifier of the assessment host. Different salts yield unlinkable identifiers. Implementations SHOULD use a per-deployment salt and MUST NOT include raw hardware identifiers. Artifact paths in model_digest can reveal project structure, so publishers MAY replace them with opaque labels if the digests remain stable. 10. IANA Considerations 10.1. EAT Profile Identifier This document uses the profile identifier "tag:ai-id- engine,2026:presilicon-evidence/v1". No IANA action is requested for the identifier itself. Watts Expires 20 May 2027 [Page 10] Internet-Draft Pre-Silicon Evidence Profile November 2026 10.2. Claims If this document progresses, it would request registration of a "presilicon-evidence" claim in the CBOR Web Token (CWT) Claims and JSON Web Token Claims registries, with the semantics of Section 5. Until then, implementations use a private claim key. 11. References 11.1. Normative References [FIPS180-4] National Institute of Standards and Technology, "Secure Hash Standard (SHS)", FIPS PUB 180-4, August 2015, . [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 1997, . [RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174, May 2017, . [RFC8259] Bray, T., Ed., "The JavaScript Object Notation (JSON) Data Interchange Format", STD 90, RFC 8259, DOI 10.17487/RFC8259, December 2017, . [RFC9334] Birkholz, H., Thaler, D., Richardson, M., Smith, N., and W. Pan, "Remote ATtestation procedureS (RATS) Architecture", RFC 9334, DOI 10.17487/RFC9334, January 2023, . [RFC9711] Lundblade, L., Mandyam, G., O'Donoghue, J., and C. Wallace, "The Entity Attestation Token (EAT)", RFC 9711, DOI 10.17487/RFC9711, April 2025, . 11.2. Informative References [TPM2] Trusted Computing Group, "Trusted Platform Module Library Specification, Family "2.0"", 2019, . Watts Expires 20 May 2027 [Page 11] Internet-Draft Pre-Silicon Evidence Profile November 2026 [TVLA] Goodwill, G., Jun, B., Jaffe, J., and P. Rohatgi, "A testing methodology for side-channel resistance validation", NIST Non-Invasive Attack Testing Workshop, September 2011. Implementation Status [RFC Editor: remove before publication.] The ai-id-engine project implements this profile in JSON form: * "aiid evidence create" and "aiid evidence verify" produce and check reports; * TPM 2.0 signing goes through tpm2-tools, with a software-key fallback; * every TVLA run in continuous integration produces a report with a software root; * unit tests cover tampering with behavior claims, upgrading the provenance after signing, missing provenance, wrong classes and duplicate fields. The EAT and COSE encoding of Section 5 is not yet implemented. Author's Address Deonte Watts ai-id-engine project Email: deonte@goodshyt.fun URI: https://github.com/DeontewattsV1/ai-id-engine Watts Expires 20 May 2027 [Page 12]