Position: Build Agent Advocates, Not Platform Agents

Sayash Kapoor, Noam Kolt, Seth Lazar

TL;DR: The AI community should take steps to increase user autonomy by championing user-centric agent advocates.

Abstract: Language model agents are poised to mediate how people navigate and act online. If the companies that already dominate internet search, communication, and commerce—or the firms trying to unseat them—control these agents, the resulting *platform agents* will likely deepen surveillance, tighten lock-in, and further entrench incumbents. To resist that trajectory, this position paper argues that we should promote *agent advocates*: user-controlled agents that safeguard individual autonomy and choice. Doing so demands three coordinated moves: broad public access to both compute and capable AI models that are not platform-owned, open interoperability and safety standards, and market regulation that prevents platforms from foreclosing competition.

Lay Summary: Many companies are developing AI agents to browse the web and act on users' behalf. If existing platform companies control these assistants, they'll likely use them to reduce privacy and make it harder to switch to competitors. This paper proposes an alternative: "agent advocates"—AI agents that users control and that work for users' benefit. To make this possible, we need to ensure public access to the computing power required to develop and run agents, standards that allow interoperability as well as safety, and market regulation that increases competition.

Primary Area: Social, Ethical, and Environmental Impacts

Keywords: agents, platform agents, agent advocates, interoperability, competitiveness

Submission Number: 33