Thought Virus: Viral Misalignment via Subliminal Prompting in Multi-Agent Systems
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912931765026816 |
|---|---|
| author | Weckbecker, Moritz Müller, Jonas Hagag, Ben Mulet, Michael |
| author_facet | Weckbecker, Moritz Müller, Jonas Hagag, Ben Mulet, Michael |
| contents | Subliminal prompting is a phenomenon in which language models are biased towards certain concepts or traits through prompting with semantically unrelated tokens. While prior work has examined subliminal prompting in user-LLM interactions, potential bias transfer in multi-agent systems and its associated security implications remain unexplored. In this work, we show that a single subliminally prompted agent can spread a weakening but persisting bias throughout its entire network. We measure this phenomenon across 6 agents using two different topologies, observing that the transferred concept maintains an elevated response rate throughout the network. To exemplify potential misalignment risks, we assess network performance on multiple-choice TruthfulQA, showing that subliminal prompting of a single agent may degrade the truthfulness of other agents. Our findings reveal that subliminal prompting introduces a new attack vector in multi-agent security, with implications for the alignment of such systems. The implementation of all experiments is publicly available at https://github.com/Multi-Agent-Security-Initiative/thought_virus . |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_00131 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Thought Virus: Viral Misalignment via Subliminal Prompting in Multi-Agent Systems Weckbecker, Moritz Müller, Jonas Hagag, Ben Mulet, Michael Multiagent Systems Artificial Intelligence Subliminal prompting is a phenomenon in which language models are biased towards certain concepts or traits through prompting with semantically unrelated tokens. While prior work has examined subliminal prompting in user-LLM interactions, potential bias transfer in multi-agent systems and its associated security implications remain unexplored. In this work, we show that a single subliminally prompted agent can spread a weakening but persisting bias throughout its entire network. We measure this phenomenon across 6 agents using two different topologies, observing that the transferred concept maintains an elevated response rate throughout the network. To exemplify potential misalignment risks, we assess network performance on multiple-choice TruthfulQA, showing that subliminal prompting of a single agent may degrade the truthfulness of other agents. Our findings reveal that subliminal prompting introduces a new attack vector in multi-agent security, with implications for the alignment of such systems. The implementation of all experiments is publicly available at https://github.com/Multi-Agent-Security-Initiative/thought_virus . |
| title | Thought Virus: Viral Misalignment via Subliminal Prompting in Multi-Agent Systems |
| topic | Multiagent Systems Artificial Intelligence |
| url | https://arxiv.org/abs/2603.00131 |