Building Trust in Autonomous Commerce: A Verifiable Global Event Timeline and AI-Ready Fraud Intelligence Layer

Agentic commerce protocols such as AP2 and ACP define mechanisms for secure agent-initiated 
transactions but do not provide interoperable, tamper-evident auditability or verifiable temporal ordering of 
events across heterogeneous domains. As autonomous agents increasingly execute high-stakes commercial 
workflows — spanning product discovery, negotiation, authorization, and payment settlement — the 
absence of a shared, cryptographically verifiable audit substrate creates systemic vulnerabilities: event logs 
remain siloed within individual platforms, fraud labels lack immutable provenance and reproducibility, and 
AI training pipelines must rely on unverifiable ground truth that is susceptible to retroactive manipulation.
This paper addresses these gaps by proposing a verifiable global event timeline for agentic commerce, 
constructed from four core components: (1) canonical event schemas that enforce deterministic serialization 
across protocol implementations, (2) deterministic batch formation ensuring reproducible ordering without 
reliance on synchronized clocks, (3) Merkle-based append-only commitments that provide logarithmic-cost 
inclusion proofs, and (4) blockchain anchoring that establishes a tamper-evident temporal backbone across 
participating domains. Building on this timeline infrastructure, we introduce a cryptographically signed 
fraud marker that binds risk labels to anchored evidence through an unforgeable provenance chain, and a 
dataset lineage model that enables reproducible, tamper-evident AI training pipelines.
We formalize the integrity guarantees of each component, define a protocol-level specification suitable for 
adoption as an extension to existing agentic commerce standards, and present empirical results from a 
prototype implementation demonstrating: Merkle tree construction scales near-linearly, processing 50,000 
events in 47 milliseconds; end-to-end event verification completes in under 0.013 milliseconds regardless of 
batch size, enabling real-time audit at production scale; inclusion proof sizes grow logarithmically from 320 
bytes at 1,000 events to only 512 bytes at 50,000 events; and Merkle-based verification outperforms linear 
scan by 14.4x at 50,000 events. This work establishes a missing transparency layer for autonomous 
commerce and provides a foundation for auditable, trustworthy AI-driven economic systems.