In-Context Defense in Computer Agents: An Empirical Study
By: Pei Yang, Hai Ci, Mike Zheng Shou
Potential Business Impact:
Protects smart computers from tricky fake instructions.
Computer agents powered by vision-language models (VLMs) have significantly advanced human-computer interaction, enabling users to perform complex tasks through natural language instructions. However, these agents are vulnerable to context deception attacks, an emerging threat where adversaries embed misleading content into the agent's operational environment, such as a pop-up window containing deceptive instructions. Existing defenses, such as instructing agents to ignore deceptive elements, have proven largely ineffective. As the first systematic study on protecting computer agents, we introduce textbf{in-context defense}, leveraging in-context learning and chain-of-thought (CoT) reasoning to counter such attacks. Our approach involves augmenting the agent's context with a small set of carefully curated exemplars containing both malicious environments and corresponding defensive responses. These exemplars guide the agent to first perform explicit defensive reasoning before action planning, reducing susceptibility to deceptive attacks. Experiments demonstrate the effectiveness of our method, reducing attack success rates by 91.2% on pop-up window attacks, 74.6% on average on environment injection attacks, while achieving 100% successful defenses against distracting advertisements. Our findings highlight that (1) defensive reasoning must precede action planning for optimal performance, and (2) a minimal number of exemplars (fewer than three) is sufficient to induce an agent's defensive behavior.
Similar Papers
Defend LLMs Through Self-Consciousness
Artificial Intelligence
Keeps AI from being tricked by bad instructions.
Invasive Context Engineering to Control Large Language Models
Artificial Intelligence
Keeps AI from being tricked, even with long talks.
ExplainableGuard: Interpretable Adversarial Defense for Large Language Models Using Chain-of-Thought Reasoning
Cryptography and Security
Explains how AI is tricked and fixes it.