Verification register Agentic Commerce & Payments
Current reading
The readiness gap, in one scan
AI-assisted assembly · derived results
- Claimed
- 70
- Reported
- 63
- Verified
- 55
- Gap
- +15
Public ambition and stated capability
Observed practitioner reporting
Independently supported evidence
Claimed minus verified
Evidence strength Critical
Decision
What the current evidence supports
Human editorial judgment · 2026-06-28
Track; not yet
- Why
- Best-in-class full-duplex naturalness (0.205s avg latency, 95% interruption takeover) is genuinely useful for a voice layer, but its 19% speech-reasoning, English-only scope, GPU self-host burden, and absent independent MOS/WER make it wrong as an end-to-end agent today.
- Next
- Prototype PersonaPlex as a voice front-end (low-latency turn-taking + persona) decoupled from a separate reasoning LLM brain; benchmark interruption handling against gpt-realtime in our checkout flow before any commitment.
Constraints
Blockers
No named blocker is present in the current public projection.
Evidence summary
Derived counts
AI-assisted assembly
- Total
- 7
- Tier 1
- 0
- Tier 2
- 3
- Tier 3
- 4
- Supports
- 3
- Contradicts
- 3
- Context
- 1
- Latest observed
- 2026-06-25
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.