Throttling Web Agents Using Reasoning Gates

2025-09-01

AI Summary

AI web agents pose significant security threats—including resource exhaustion, CAPTCHA bypass, and authentication flooding—due to their automated, high-throughput interactions. Method: This paper proposes the Web Agent Throttling Framework, centered on a puzzle-based multi-hop reasoning gate. It employs an asymmetric verification protocol where puzzle generation cost substantially exceeds verification cost, imposing tunable computational overhead on malicious or misconfigured agents. The framework is client-agnostic, compatible with diverse LLM-based agents, and supports both synthetic text puzzles and language-model-driven reasoning challenges. It integrates with MCP servers and custom websites for end-to-end deployment. Contribution/Results: Experiments demonstrate that the framework increases response generation cost by 9.2× over state-of-the-art models, effectively mitigating automated abuse while preserving service availability and demonstrating practical deployability.

Technology Category

• Multiagent Systems: Adversarial Agents
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Natural Language Processing: Safety and Robustness

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Economics, Online Markets and Human Computation: Sustainability of Web economics

Abstract

AI web agents use Internet resources at far greater speed, scale, and complexity -- changing how users and services interact. Deployed maliciously or erroneously, these agents could overload content providers. At the same time, web agents can bypass CAPTCHAs and other defenses by mimicking user behavior or flood authentication systems with fake accounts. Yet providers must protect their services and content from denial-of-service attacks and scraping by web agents. In this paper, we design a framework that imposes tunable costs on agents before providing access to resources; we call this Web Agent Throttling. We start by formalizing Throttling Gates as challenges issued to an agent that are asymmetric, scalable, robust, and compatible with any agent. Focusing on a common component -- the language model -- we require the agent to solve reasoning puzzles, thereby incurring excessive token-generation costs. However, we find that using existing puzzles, e.g., coding or math, as throttling gates fails to satisfy our properties. To address this, we introduce rebus-based Reasoning Gates, synthetic text puzzles that require multi-hop reasoning over world knowledge (thereby throttling an agent's model). We design a scalable generation and verification protocol for such reasoning gates. Our framework achieves computational asymmetry, i.e., the response-generation cost is 9.2x higher than the generation cost for SOTA models. We further deploy reasoning gates on a custom website and Model Context Protocol (MCP) servers and evaluate with real-world web agents. Finally, we discuss the limitations and environmental impact of real-world deployment of our framework.

Problem

• Throttling web agents to prevent service overload
• Imposing asymmetric costs on agents via reasoning puzzles
• Protecting web services from denial-of-service attacks

Innovation

• Rebus-based reasoning gates for throttling
• Asymmetric computational costs on agents
• Scalable generation and verification protocol