Verification register Security & Identity
Current reading
The readiness gap, in one scan
AI-assisted assembly · derived results
- Claimed
- 60
- Reported
- 55
- Verified
- 49
- Gap
- +11
Public ambition and stated capability
Observed practitioner reporting
Independently supported evidence
Claimed minus verified
Evidence strength Growing
Decision
What the current evidence supports
Human editorial judgment · 2026-06-28
Track; not yet
- Why
- Primary-vendor, open-source, and conceptually aligned with pv's 'expose value not records' posture, but brand-new, preview-coupled, and unproven in our stack; the attested-TLS pattern is promising but needs our own measurement and a threat-model review of what the handshake does NOT measure.
- Next
- Clone github.com/google/prompt-encryption-sdk, run the Codelab against a Confidential G4 preview VM, and measure attested-TLS handshake + inference latency overhead on a representative pv inference call (realistic prompt sizes, small + large batch) before any production wiring; pair with a TEE threat-model review against the Trail of Bits unmeasured-input findings.
Constraints
Blockers
No named blocker is present in the current public projection.
Evidence summary
Derived counts
AI-assisted assembly
- Total
- 8
- Tier 1
- 0
- Tier 2
- 2
- Tier 3
- 6
- Supports
- 2
- Contradicts
- 3
- Context
- 3
- Latest observed
- 2026-06-10
Counts and dates only. Raw signals, private excerpts, trust records, and internal corpus material are not published here.
Publication record
Revisions
Initial public reading
This is the initial public reading. No earlier readiness change is recorded.