A Trembling House of Cards? Mapping Adversarial Attacks against Language Agents

arXiv:2402.10196v1 Announce Type: new
Abstract: Language agents powered by large language models (LLMs) have seen exploding development. Their capability of using language as a vehicle for thought and communication lends an incredible level of flexibility and versatility. People have quickly capitalized on this capability to connect LLMs to a wide range of external components and environments: databases, tools, the Internet, robotic embodiment, etc. Many believe an unprecedentedly powerful automation technology is emerging. However, new automation technologies come with new safety …

abstract adversarial adversarial attacks agents arxiv attacks capability cards communication cs.ai cs.cl development flexibility language language models large language large language models llms mapping people thought type