Build the web for agents, not agents for the web

2025-06-12

AI Summary

Current web agents rely on human-centric interfaces—such as complex DOM structures, screenshots, or API bypasses—that fundamentally mismatch the capabilities of LLMs and multimodal models, resulting in low efficiency and poor reliability. To address this, we propose a paradigm shift: the Agentic Web Interface (AWI), a web interaction paradigm explicitly optimized for autonomous agents. We systematically define AWI and introduce six design principles centered on agent-first priorities—security, efficiency, and standardization. Our methodology integrates LLM and multimodal perception, human-computer interaction theory, web standards (e.g., W3C specifications), and trustworthy AI principles. Our key contribution is the first agent-native web interface framework, providing foundational support for efficient, reliable, and interpretable web agents. Furthermore, AWI lays the groundwork for a new ecosystem of agent-oriented web protocols, fostering collaborative development across the ML community. (149 words)

Technology Category

• Humans and AI: Intelligent User Interfaces
• Multiagent Systems: Agent Communication
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Agentic search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Recent advancements in Large Language Models (LLMs) and multimodal counterparts have spurred significant interest in developing web agents -- AI systems capable of autonomously navigating and completing tasks within web environments. While holding tremendous promise for automating complex web interactions, current approaches face substantial challenges due to the fundamental mismatch between human-designed interfaces and LLM capabilities. Current methods struggle with the inherent complexity of web inputs, whether processing massive DOM trees, relying on screenshots augmented with additional information, or bypassing the user interface entirely through API interactions. This position paper advocates for a paradigm shift in web agent research: rather than forcing web agents to adapt to interfaces designed for humans, we should develop a new interaction paradigm specifically optimized for agentic capabilities. To this end, we introduce the concept of an Agentic Web Interface (AWI), an interface specifically designed for agents to navigate a website. We establish six guiding principles for AWI design, emphasizing safety, efficiency, and standardization, to account for the interests of all primary stakeholders. This reframing aims to overcome fundamental limitations of existing interfaces, paving the way for more efficient, reliable, and transparent web agent design, which will be a collaborative effort involving the broader ML community.

Problem

• Mismatch between human interfaces and LLM capabilities
• Challenges in processing complex web inputs efficiently
• Need for agent-specific web interfaces (AWI) optimization

Innovation

• Agentic Web Interface for AI navigation
• Six principles for agent-friendly design
• Shift from human to agent interfaces