Protocol-Level Attacks on Agentic Commerce Platforms: A Cross-Platform Taxonomy, AIP-Bench, and Unified Defense

Abstract

Agentic commerce platforms let AI agents autonomously discover services, move payments, and wield user credentials on their users' behalf, and they already handle real money. Their security has so far been studied almost entirely at the level of the AI model, through prompt injection and misalignment. We show that the more consequential risks lie one layer down, in the protocol between agents and commerce services. There, vulnerabilities are structural : exploitation is deterministic and ndepend

Categories

Framework mappings

Suggested from the entry's categories.

@misc{louck2026protocollevel,
  title = {{Protocol-Level Attacks on Agentic Commerce Platforms: A Cross-Platform Taxonomy, AIP-Bench, and Unified Defense}},
  author = {Yedidel Louck},
  year = {2026},
  month = jul,
  eprint = {2607.21824},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2607.21824}
}