Smarter Saboteurs, Better Fixers: Scaling & Security in Linear Multi-Agent Workflows

June 10, 2026 Β· Grace Period Β· πŸ› the AIWILD Workshop at ICML 2026

⏳ Grace Period
This paper is less than 90 days old. We give authors time to release their code before passing judgment.
Authors Timothy McAllister, Sina Abdidizaji, Ivan Garibay, Ozlem Ozmen Garibay arXiv ID 2606.12709 Category cs.MA: Multiagent Systems Cross-listed cs.CR, cs.LG Citations 0 Venue the AIWILD Workshop at ICML 2026
Abstract
As LLM-based multi-agent systems (MAS) are deployed in the wild, the resilience of their collaboration structures against adversarial compromise becomes a critical safety concern. Attackers may leverage prompt-injection or jailbreaking to sabotage individual agents within MAS workflows, but the interaction between model scaling and system-level resilience remains poorly understood. This paper investigates how model scale affects the security of linear multi-agent workflows. Our experiments across scales of two open-weight model families on the HumanEval benchmark reveal a compliance-correction symmetry: larger models are far more likely to faithfully execute malicious instructions, with the control-to-malicious performance drop reaching 53.7pp at 27B in uncorrected pipelines. However, appending a lightweight terminal Fixer stage collapses this to 0.6pp and restores statistical parity with control-level performance, demonstrating that strictly linear collaboration structures can be viable and resilient to adversaries at this scale, and suggesting that the brittleness previously attributed to linear topology may stem from a lack of correction.
Community shame:
Not yet rated
Community Contributions

Found the code? Know the venue? Think something is wrong? Let us know!

πŸ“œ Similar Papers

In the same crypt β€” Multiagent Systems