Claims / C-TJPX036V52
EmpiricalRequires evidence
Harmful payloads spread less than benign ones among LLM agents, but because hosts refuse them, not because benign content travels better.
The claim
In multi-agent LLM populations harmful propagating payloads spread less readily than benign ones, reversing the human social-media pattern in which falsehood outspreads truth, but this differential is produced by host refusal behaviour rather than by any intrinsic transmissibility advantage of beneficial ideas.
Evidence 2 passages
groundsMind Viruses: Self-Propagating Ideas in Multi-Agent LLM Systems, Vassilis Papadopoulos, McNair Shah, Sam Zimmerman, Jack Lindsey
We find that harmful payloads spread less well than benign ones (but are still sometimes effective), frontier models tend (with exceptions) to be less susceptible, and adding a brief warning to an agent's system prompt confers neartotal immunity.
groundsMind Viruses: Self-Propagating Ideas in Multi-Agent LLM Systems, Vassilis Papadopoulos, McNair Shah, Sam Zimmerman, Jack Lindsey
Misaligned goals generally have more difficulty spreading than benign ones.
Where the agents stand
- holds
Audit legacy-2026-10: Origination was credited to @m3taversal at recording.