P126 Independent readiness review

Have the mechanisms that verification proposals depend on had an independent technology-readiness review, and do independent ratings match the proponents' own?

Whether the mechanisms verification proposals depend on have had an independent technology-readiness review, whether independent ratings match the proponents' own, and the deployment timelines the sources give per mechanism family.

Status

Analyzed A literature-based feasibility taxonomy of about twenty hardware-governance mechanisms by an author outside the proponent groups; by its own limitations section the ratings come from the author's reading 'rather than from formal expert elicitation or independent technology readiness review'; no comparison of independent ratings against proponents' own. The per-family timelines in the sources (Cankaya: off-chip mechanisms in testing and de-risking as of July 2026; Harack et al.: a one-year crash program; Aarne et al.: about four years for on-chip uptake) are proponents' estimates and belong to those families' problems.

Our analysis reveals a structural mismatch: the mechanisms most needed for treaty verification … are also the least mature.
Ansari (2026), Hardware-Level Governance of AI Compute: A Feasibility Taxonomy for Regulatory Compliance and Treaty Verification (arXiv 2604.04712v1) (2026-04-06), L30–32

Raised in: Harack et al. (2025), Verification for International AI Governance (Oxford AIGI); Petrie, Aarne, Ammann & Dalrymple (2025), Overview of Flexible Hardware-Enabled Guarantees (flexHEG Part 1); Petrie & Aarne (2025), Technical Options for Flexible Hardware-Enabled Guarantees (flexHEG Part 2); Aarne, Fist & Withers (2024), Secure, Governable Chips (CNAS); Scher (2026), Verifying Restrictions on Frontier AI Research (MIRI); Cankaya (2026), All Hands on Deck to Build the Datacenter Lie Detector; Cankaya (2026), Compute Verification FAQs; Canary Institute research; Ansari (2026), Hardware-Level Governance of AI Compute: A Feasibility Taxonomy for Regulatory Compliance and Treaty Verification; Harack (2026), AI Verification: Infrastructure for Prosperity, Governance, and Peace (Lawfare); Koopmanschap & Barten (2026), How to Catch a GPU: A Taxonomy of Verification and Enforcement Mechanisms for International AI Agreements; AI 2040 (2026), Get Involved in Verification; Dean, Verification Plan (AI 2040); Proofworks, Living Verification Documents.

Sub-questions

IDSub-questionWhat the sources report
P126.1How long will retrofittable off-chip mechanisms take to deploy at scale, given they are (as of July 2026) still in testing and de-risking?open
P126.2Has an end-to-end network-tap + recomputation inference-only retrofit been demonstrated on a production-scale rack?open
P126.3If a deal were signed now, could it be enforced trustlessly at low cost, and if not, what fraction of frontier AI services would need to stay offline and for how long?open
P126.4When will on-chip / flexHEG mechanisms with high physical security be deployed widely enough to be effective (source targets: at scale by 2027; ~4 years uptake after hardening)?open
P126.5How long would a crash program take to make a few key data centres verifiable (~1 year per source) and a broader ecosystem (a few years per source)?a source suggests an answer (not checked)
P126.6Do the mechanisms treaty proposals depend on (on-chip metering / chip-use verification, proof of training, hardware enforcement, privacy-preserving verification) all remain below governance grade (low on the paper's four-point feasibility scale)?open
P126.7Do an independent expert panel's TRL ratings for the 20 mechanisms match the paper's own four-tier feasibility ratings?open

Sub-questions not yet stated precisely enough to check

IDSub-question
P126.8Are off-chip components (network taps, analog sensors) technically feasible at scale against top-tier adversaries?
P126.9What is the actual readiness of unpublished work (the MVP consensus whose specifics are TBA; redacted projects)?
P126.10Have the proposed solutions to prevent-escape and prevent-acquisition been finalised, with implementation details and reduced cost?
P126.11Has prior work said how to verifiably regulate AI development and deployment?

Possibly relevant work found by automated search

Not searched: no technical result was expected for this problem.

Data