Allocation Readiness Level (ARL) Scorecard

How ready are today's agentic payment stacks to answer who deserves the value? Based on the public design constraints in The AI Attribution Trap (R, 2026).
⚠️ This is a structural assessment of published architecture, not a vendor audit. Scores reflect what each stack publicly discloses about allocation design — not product quality, security, or commercial viability. All data from public documentation, Aug 2026. Corrections welcome: we invite counterexamples.
Median ARL: 1
0 of 8 stacks reach ARL 3+. Allocation is the missing layer across the board.
C3 Failures: 8/8
No stack publicly decouples measurement weight from capital power.
Stack ARL What exists Missing / Fails Basis
x402 Foundation
Agentic payment protocol		 1 Verified payment rail; request–response settlement; wallet controls. C1 No verifiable contribution grounds
C2 No multi-party allocation
C3 Capital-aligned governance
C4 No linkage to realized value		 Public docs + Chainalysis report
Circle Agent Stack
Operating layer for agents		 1 Agent identity, wallet infra, policy controls. C1 No contribution verification
C3 Governance not disclosed
C4 Points/credits risk		 Circle launch docs (2026-05/07)
Google AP2 + x402
Agentic payments protocol		 1 Intent-based payment protocol; HTTP-native. C1 Attribution external to protocol
C2 Single-payer model
C3 n/a disclosed		 Google 2025-09 announcement
OpenAI Agent Payments
Ecosystem payment		 1 Agent wallet; usage-based billing. C1 No contribution measurement
C3 Platform-controlled
C4 Internal credits		 Public API docs
Anthropic Agent SDK + Claude
Agent framework		 0 Agent runtime; no payment/allocation layer. C1-C4 Allocation entirely out of scope SDK public docs
Stripe Agent Payments
Payment infra for agents		 1 Payment rails; merchant-of-record. C1 Transaction = contribution? No
C3 Centralized
C4 Value linkage implicit		 Public docs
Agent wallets (Metamask/Coinbase)
Agent custody		 0 Custody; key management; transaction signing. C1-C4 No allocation semantics Product docs
Base chain agentic volume
Settlement layer		 1 100M+ agentic transactions; settlement finality. C1 Transactions ≠ contributions
C2 No multi-party allocation
C3 Validator/capital control		 Chainalysis (2026)
The AI Attribution Trap · R (independent researcher) · ARL 0–5 scale defined in §6.1 of the paper · Invites counterexamples — if your stack answers C1–C4, tell us. · Zero proprietary logic.