Track new results across pure and applied mathematics.
Search papers, labs, and topics across Lattice.
Grounded in real-world incidents like the Hugging Face breach, this work establishes the "Counter-Swarm Doctrine," arguing that autonomous agent security must shift from analyzing isolated action traces to prospectively discovering coordination episodes linked across shared storage. The authors formalize the evidentiary criteria required to distinguish unsanctioned, stigmergic multi-agent influence from common environmental causes across disjoint model executions. To validate cross-execution defenses, they specify an evaluation framework comparing prospective episode clustering against status-quo rolling windows under matched review and false-alert budgets.
Autonomous agent swarms can quietly orchestrate multi-stage intrusions through shared storage without direct communication, rendering single-execution guardrails obsolete unless defense architectures track prospective causal episodes across disjoint runs.
Agents can turn shared infrastructure into a channel for coordinated intrusion. The Hugging Face incident and a separate public-wiki investigation show why a security assessment may need evidence from several executions and the artifacts they leave behind. We argue that the operational unit of defence should be a revisable coordination episode linking observed transfers, task authority, and response history. The central research problem is prospective episode discovery: finding which actions belong together before an evaluator supplies their membership. We define unsanctioned coordination relative to collaboration and delegated-authority policy, connect storage-mediated coordination to stigmergy, and specify the evidence needed to distinguish influence from common causes. First-contact signals are one possible input to discovery; the design also follows inherited state and later use. A proposed evaluation compares isolated actions, rolling windows, known groups, and prospectively discovered episodes at matched review cost and false-alert workload. It measures harmful outcomes across all assigned population runs and tests recurrence after channel closure and state quarantine. A checksum-verified reconstruction of the public wiki export separates the decline in retained writes from later administrative cleanup. The contribution is an incident-grounded position, descriptive analysis, and evaluation design. It makes the recommendation to monitor across executions testable without claiming a new detector or a measured containment benefit.