AI Power Concentration Observatory

1. What is it

The AI Concentration of Power Observatory tracks early signs of AI-driven concentration of power and serves as a reference point for researchers, journalists, policymakers, and civil society actors responding to emerging risks.

2. Scope of Version 1

In Version 1, I will focus on three signaling areas: AI-mediated influence and epistemic infrastructure, the integration of AI into decision-making in government and institutions, and unequal access to advanced AI capabilities. I begin with epistemic infrastructure because it is a neglected yet increasingly important way AI systems can shape the information environment, public understanding, and collective decision-making, and it directly relates to my work as part of the Corda Democracy Fellowship. The broader monitoring framework will also include additional avenues, such as government reliance on advanced AI providers and governance changes within AI companies, which can be expanded upon after testing the initial methodology.

3. Who is involved + Authority

Initially, I will lead the project myself. I have already begun building the Observatory and will leverage my experience with AI governance and coordination projects gained through the Corda Democracy Fellowship, AI Safety Camp, and Apart Research, including stakeholder engagement, research coordination, and the development of early-stage ideas into working prototypes.

4. Specific Short-Term Deliverables

With this grant, I will deliver a working version 1 of the Observatory, covering selected signaling areas, a structured and regularly updated database of relevant events, a publicly available monitoring system with indicators, and the first published analysis identifying emerging patterns and risks.

5. Method + Feedback

I will collect and analyze information from open sources, including policy documents, government statements, company announcements, legislative changes, media investigations, and civil society research. I will test the methodology directly with potential users—including politicians, journalists, researchers, think tanks, and civil society organizations—and continually refine the framework based on their feedback.

6. Why it matters

By making early signs of AI-enabled power concentration more visible and easier to understand, the Observatory aims to help researchers, journalists, policymakers, and civil society actors identify concerning developments early, while there is still time to investigate, increase transparency, strengthen accountability mechanisms, and question, challenge, or redirect these trends.