# AI verification proposal

A proposal built with the Proposal Explorer of the AI Verification Tech Map (https://trustbutveri.fyi/), from its records of 2026-10-09. Interactive version: https://trustbutveri.fyi/explorer/?mechanisms=M-0012,M-0014&implementations=M-0014:I-0010

How to read it: a claim is something one party wants to verify about another's AI hardware or software. A mechanism is a general technique for verifying claims; it is "aimed at" a claim when that is its direct purpose, and "supporting" when it contributes without being aimed at it. A claim is addressed when a mechanism in the proposal is aimed at it and is not excluded by the filters; addressed does not mean verified, so check that mechanism's development status, security evidence and findings. Definitions: https://trustbutveri.fyi/about/methodology/ (roles, properties and findings) and https://trustbutveri.fyi/about/readiness/ (development status).

## Filters

Filters apply to mechanisms only and describe the setting the proposal is for.

None set. Every mechanism on the map was available.

## Overview

One row per mechanism, read from its record. Open failures: critical / significant / minor. The last three columns are the editors' reading of what the verifier sees. Findings are grouped as known failures, scope limitations and open questions. Only known failures count as failures. Counts are an inventory of published findings, not a risk score.

| Mechanism | Development | Security evidence | Prover | Attack testing | Hardware | Open failures | Weights | Inputs and outputs | Training data |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Hardware-attested weight binding | Operational use | Published attack testing | Semi-trusted | Independent red-team | Existing features | 1 / 0 / 0 | hidden | unspecified | not involved |
| Bandwidth limits and compartmentalization / RAND secure inference data center (SIDC) design | Proposed | Published security analysis | Semi-trusted | Analysis | Retrofit device | 0 / 0 / 0 | unspecified | unspecified | unspecified |

## Claims

No claims chosen.

## Mechanisms

### Hardware-attested weight binding

Checks that a hardware enclave serves committed model weights, by attesting software that tests the weights against a hash commitment when they are read. ([Hardware-attested weight binding](https://trustbutveri.fyi/mechanisms/model-identity-attestation/))

- Assessment: mechanism family.
- Development: Operational use (legacy code R3), assessed for hardware-attested weight binding showing users that a service runs its committed weights.
- Security evidence: Published attack testing. Independent evaluation: unassessed. Formal proof: unassessed. Deployment assurance: unassessed.
- Claims in this proposal: none of them.
- Threat model: semi-trusted prover. Hardware: existing features. Prover cooperation: required. Attack testing: independent red-team. Category: Cryptographic & computational.
- What the verifier sees: model weights hidden; inputs and outputs unspecified; training data not involved. Verifiers check a hash commitment to the weights, carried in a hardware attestation, and need no access to the weights themselves. The mechanism does not specify whether prompts and outputs are disclosed to a verifier. It involves no training data.

### Bandwidth limits and compartmentalization

A RAND design for a purpose-built facility that serves already-trained AI models while protecting weights and inference data against state-level attackers. ([Bandwidth limits and compartmentalization](https://trustbutveri.fyi/mechanisms/bandwidth-limits-and-compartmentalization/))

- Assessment: selected implementation [RAND secure inference data center (SIDC) design](https://trustbutveri.fyi/implementations/rand-secure-inference-data-centers/).
- Development: Proposed (legacy code R1), assessed for the operator's own weight security, with no outside verification described.
- Security evidence: Published security analysis. Independent evaluation: unassessed. Formal proof: unassessed. Deployment assurance: unassessed.
- Claims in this proposal: none of them.
- Threat model: semi-trusted prover. Hardware: retrofit device. Prover cooperation: required. Attack testing: analysis. Category: Isolation & system architectures.
- What the verifier sees: model weights unspecified; inputs and outputs unspecified; training data unspecified. This Explorer has no asset-specific exposure assessment for this implementation. Check its source and deployment assumptions.


## Properties

**Operational use**

- Hardware-attested weight binding: Operational use (legacy code R3), assessed for hardware-attested weight binding showing users that a service runs its committed weights

**No new hardware needed**

- Hardware-attested weight binding

**Failures since mitigated**

- Launch-state attestation does not by itself cover weights loaded later (in Hardware-attested weight binding) [1][11]


## Attack testing

Attack testing records published testing for this use. It does not by itself show independent review, a formal proof or that a deployed system is secure.

**Testing history**

- Hardware-attested weight binding: Independent red-team
- Bandwidth limits and compartmentalization / RAND secure inference data center (SIDC) design: Analysis


## Limits

**Open critical failures**

- Underlying attestation can be forged or relayed (known failure, demonstrated attack, in Hardware-attested weight binding; https://trustbutveri.fyi/mechanisms/model-identity-attestation/evidence/flaws/1/) [2][3][7][8][9][10]. Inherited finding. Critical for weight binding against an operator with physical access to affected hardware, or with control of the hypervisor on an AMD SEV-SNP platform without AMD's fixes. PAL*M excludes physical attacks, and Tinfoil acknowledges this boundary. The enclave route inherits the platform-specific TEE attestation failures. Intel TDX forgery and H100 relay were demonstrated with physical access and host control. Battering RAM defeated AMD SEV-SNP attestation on DDR4 servers; RMPocalypse did so from malicious host software on platforms without AMD's fixes. These demonstrate failures of the trust roots, not of each model-commitment protocol. Related finding: https://trustbutveri.fyi/mechanisms/tee-remote-attestation/evidence/flaws/1/.

  Response: The TEE.fail authors report that physical interposer attacks are outside Intel's and AMD's threat models. AMD reports fixes for RMPocalypse.

  Related mechanism: Hardware-enabled guarantees (flexHEG) and guarantee processors (R1, not in the proposal). A tamper-protected enclosure around the chip is the proposed answer when the party that holds the hardware may attack it physically.

**Family finding context**

- Context for RAND secure inference data center (SIDC) design. Findings from the mechanism family appear here as context. They apply to an implementation only when its own record lists them, under the conditions stated there. Low-communication training reduces the bandwidth training needs (open question, theoretical argument, in Bandwidth limits and compartmentalization; https://trustbutveri.fyi/mechanisms/bandwidth-limits-and-compartmentalization/evidence/flaws/1/) [14][15][16]. DiLoCo matched fully synchronous training on 8 workers while communicating 500 times less. Rahman writes that this family of methods theoretically allows large-scale training with less than 100 Mbps. Lucid includes these methods in its bounds, but notes that extreme activation compression, architectures with unusually small inter-layer widths, or modular paradigms could erode the margin.
- Context for RAND secure inference data center (SIDC) design. Findings from the mechanism family appear here as context. They apply to an implementation only when its own record lists them, under the conditions stated there. Operator control of pod routing collapses the bound (known failure, theoretical argument, in Bandwidth limits and compartmentalization; https://trustbutveri.fyi/mechanisms/bandwidth-limits-and-compartmentalization/evidence/flaws/2/) [16]. Lucid's analysis finds that if the operator can freely assign pods to routers, it could dedicate a whole cell of 100 or more pods to one pipeline stage. The bound then falls to about 90–220x uncompressed and as low as about 25x with compression. The proposed mitigation, auditor-controlled random assignment that is periodically re-randomized, has not been implemented.
- Context for RAND secure inference data center (SIDC) design. Findings from the mechanism family appear here as context. They apply to an implementation only when its own record lists them, under the conditions stated there. Undeclared local storage raises per-pod capacity (scope limitation, theoretical argument, in Bandwidth limits and compartmentalization; https://trustbutveri.fyi/mechanisms/bandwidth-limits-and-compartmentalization/evidence/flaws/3/) [16]. More memory or storage per pod helps an adversary. Lucid requires per-pod storage to be declared, capped and physically inspected.
- Context for RAND secure inference data center (SIDC) design. Findings from the mechanism family appear here as context. They apply to an implementation only when its own record lists them, under the conditions stated there. Training within one pod is not covered (scope limitation, open question, in Bandwidth limits and compartmentalization; https://trustbutveri.fyi/mechanisms/bandwidth-limits-and-compartmentalization/evidence/flaws/4/) [16]. Lucid's bounds concern pre-training models larger than the pods are sized for. Training models that fit in one pod, fine-tuning and reinforcement-learning post-training within one pod are outside the modelled threat.
- Context for RAND secure inference data center (SIDC) design. Findings from the mechanism family appear here as context. They apply to an implementation only when its own record lists them, under the conditions stated there. Parallel scale-up switches are hard enforcement points (known failure, theoretical argument, in Bandwidth limits and compartmentalization; https://trustbutveri.fyi/mechanisms/bandwidth-limits-and-compartmentalization/evidence/flaws/5/) [17]. In GB200 topologies, GPUs reach GPUs in other nodes through NVSwitches without a NIC on the path. Amodo notes that limits are hard to enforce there because many switches work in parallel, so compromising one or two would bypass the limit.

**Scope limitations**

- For private models, a user can confirm consistency but not content (scope limitation, open question, in Hardware-attested weight binding; https://trustbutveri.fyi/mechanisms/model-identity-attestation/evidence/flaws/3/) [1][12]. When weights are not published, users can check that the same root hash is served each time, but not what the model is. Pairing the hash with an attested evaluation, as in Attestable Audits, is one proposed remedy.
- Everything rests on the trusted setup (scope limitation, theoretical argument, in RAND secure inference data center (SIDC) design; https://trustbutveri.fyi/implementations/rand-secure-inference-data-centers/evidence/flaws/1/) [13]. Reference measurements for model weights and reference data are established in a trusted setup phase. The report states that the system cannot detect compromise that happened before ingestion if the trusted setup itself is compromised.
- Security weakens over long operation (scope limitation, theoretical argument, in RAND secure inference data center (SIDC) design; https://trustbutveri.fyi/implementations/rand-secure-inference-data-centers/evidence/flaws/2/) [13]. The authors claim that the facility can withstand attacks at the OC5 level for a five-year operational period. They expect its ability to withstand long OC5 campaigns to become less robust the longer the facility remains in operation.

**Not yet demonstrated**

- Bandwidth limits and compartmentalization: Proposed (legacy code R1), assessed for the operator's own weight security, with no outside verification described


## Possible additions

Mechanisms on the map, not in the proposal, that the records connect to an unaddressed or partly addressed claim, an open failure or a dependency. Pointers, not recommendations: each brings its own readiness level and findings, and none is claimed to close a failure.

- **Hardware-enabled guarantees (flexHEG) and guarantee processors** (Proposed (legacy code R1), assessed for checking and enforcing training-compute limits on chips, against adversaries up to states)
  - Bears on the open critical failure "Underlying attestation can be forged or relayed" in Hardware-attested weight binding. A tamper-protected enclosure around the chip is the proposed answer when the party that holds the hardware may attack it physically.
- **TEE remote attestation for AI workloads** (Operational use (legacy code R3), assessed for showing which software ran to a party that distrusts the operator holding the hardware)
  - Hardware-attested weight binding waits on it: Attestation that resists physical attackers, for the enclave variant.


## Dependencies

**Missing prerequisites**

- TEE remote attestation for AI workloads (Operational use (legacy code R3), assessed for showing which software ran to a party that distrusts the operator holding the hardware), needed by Hardware-attested weight binding

**Blockers**

- Hardware-attested weight binding: Attestation that resists physical attackers, for the enclave variant. (hardware trust; waits on TEE remote attestation for AI workloads) [7]
- Bandwidth limits and compartmentalization: No prototype exists; RAND recommends prototyping key security features and integration now. (adversarial validation) [13]
- Bandwidth limits and compartmentalization: The report describes internal integrity checks, audit logging and accreditation, but no way for a party outside the operator to verify the facility's properties. (access & governance) [13]
- Bandwidth limits and compartmentalization: Human review of every prompt and response makes each request take three to five minutes, with the review steps as the rate-limiting factor. (performance & compatibility) [13]
- Bandwidth limits and compartmentalization: Detailed design information is withheld from the public report and is to be evaluated privately with stakeholders, which limits independent public scrutiny. (access & governance) [13]


## What the verifier sees

- Model weights: shown by none; depends on the design for none; hidden by Hardware-attested weight binding; not involved in none; unspecified for Bandwidth limits and compartmentalization.
- Inputs and outputs: shown by none; depends on the design for none; hidden by none; not involved in none; unspecified for Hardware-attested weight binding and Bandwidth limits and compartmentalization.
- Training data: shown by none; depends on the design for none; hidden by none; not involved in Hardware-attested weight binding; unspecified for Bandwidth limits and compartmentalization.

## Implementations

- Hardware-attested weight binding: [Attestable Audits](https://trustbutveri.fyi/implementations/attestable-audits/) (R2, research prototype); [PAL*M](https://trustbutveri.fyi/implementations/palm/) (R2, research prototype); [Tinfoil model identity (Modelwrap)](https://trustbutveri.fyi/implementations/tinfoil-model-identity/) (R3, product)
- Bandwidth limits and compartmentalization: [AI 2040 inference-only verification stack](https://trustbutveri.fyi/implementations/ai-2040-inference-only-verification-plan/) (R1, proposed architecture); [RAND secure inference data center (SIDC) design](https://trustbutveri.fyi/implementations/rand-secure-inference-data-centers/) (R1, proposed architecture)

## Sources

1. How Tinfoil Proves Exactly What Model Is Running, Tinfoil Team (2026). https://tinfoil.sh/blog/2026-02-03-proving-model-identity
2. PAL*M: Property Attestation for Large Generative Models, P. Chantasantitam et al. (2026). https://arxiv.org/abs/2601.16199
3. A primer on secure enclaves, Tinfoil (2026). https://docs.tinfoil.sh/verification/secure-enclave-primer
4. Backend infrastructure, Tinfoil (2026). https://docs.tinfoil.sh/verification/attestation-architecture
5. How verification works in Tinfoil, Tinfoil (2026). https://docs.tinfoil.sh/verification/verification-in-tinfoil
6. modelwrap: Reproducible dm-verity read-only image of Huggingface models, Tinfoil (2026). https://github.com/tinfoilsh/modelwrap
7. TEE.fail: Breaking Trusted Execution Environments via DDR5 Memory Bus Interposition, J. Chuang et al. (2026). https://tee.fail/
8. Battering RAM: Low-Cost Interposer Attacks on Confidential Computing via Dynamic Memory Aliasing, J. De Meulemeester et al. (2026). https://batteringram.eu/
9. RMPocalypse: How a Catch-22 Breaks AMD SEV-SNP, B. Schlüter & S. Shinde (2025). https://rmpocalypse.github.io/
10. SEV-SNP RMP Initialization Vulnerability (AMD-SB-3020), AMD (2025). https://www.amd.com/en/resources/product-security/bulletin/amd-sb-3020.html
11. On TEEs for Privacy-Preserving Monitoring in AI Governance, Gloria Z (2026). https://techgov.intelligence.org/blog/on-tees-for-privacy-preserving-monitoring-in-ai-governance
12. Attestable Audits: Verifiable AI Safety Benchmarks Using Trusted Execution Environments, C. Schnabl et al. (2025). https://arxiv.org/abs/2506.23706
13. Highly Secure Inference Data Centers: A Vertically Integrated Strategy for Security Engineering, S. F. Comer et al. (2026). https://www.rand.org/pubs/research_reports/RRA4827-1.html
14. DiLoCo: Distributed Low-Communication Training of Language Models, A. Douillard et al. (2024). https://arxiv.org/abs/2311.08105
15. Does Distributed Training Undermine Compute Governance?, R. Rahman (2026). https://arxiv.org/abs/2605.29359
16. Traffic Shaping for Workload Classification, Lucid Computing (2026). https://lucidcomputing.substack.com/p/traffic-shaping-for-workload-classification
17. The Tray as a Bandwidth Boundary, Amodo Design (2026). https://amododesign.com/notes/2026-03-16-dpu-bandwidth-limiter/