Findings/injections-propagate-between-agents

A single malicious message can spread on its own across networks of connected agents.

Corroboratedmeasured4 evidence records from 4 independent sourcesassistant-drafted
Scope: what this does not show

Laboratory settings, one platform red-team and a September 2026 simulation of synthetic agent networks; no in-the-wild propagation recorded. In the September 2026 simulation the main results use an attacker-operated endpoint that reinserts the attack at each hop, the authors say the rates do not measure unaided reproduction, and there is no direct agent-to-agent channel; one model fully infected 21% of assistants without the endpoint.

Corroborated: Supported by at least two independent sources.

Evidence

Status history

  1. 2024-03-05ReportedMorris II: a self-replicating prompt spreads across email assistants in a lab. · record
  2. 2026-04-30CorroboratedMicrosoft red team shows propagation across a multi-agent platform. · record