Geopolitics is upstream of AI risk.

The decisions that shape frontier AI are made in ministries and summits as much as in labs. Red Line tracks that layer.

Read the founding essay

Red Line’s flagship publication, The Brief, is a rolling research agenda on the upstream geopolitics of AI risk.

Founded in 2025 by Arwen Smit, Red Line produces independent, public analysis of how power shapes the direction of AI. Smit researches open-source AI and great power competition at Cambridge University and has spent a decade at the intersection of emerging technology, policy, and risk.

Red Line also works directly with governments and institutions on scenario analysis, risk triage, and policy strategy.

Work with us · @arwengeist on X


The Brief: Archives

Anthropic said "No" to the Department of War: What does this mean for AI safety?

·
Mar 2
Anthropic said "No" to the Department of War: What does this mean for AI safety?

The Situation: This episode marks the securitisation of AI safety.

State of Open-Source AI, 2025

·
December 4, 2025
State of Open-Source AI, 2025

The Situation: Open Model 2025 Trends Are In

The Genesis Mission and the Rise of Sovereign AI

·
November 27, 2025
The Genesis Mission and the Rise of Sovereign AI

The Situation: The White House Launches the Genesis Mission EO

Stop asking bad questions on AI risk

·
November 21, 2025
Stop asking bad questions on AI risk

The Counterfactual: In AI, Some Questions Increase Risk

When China Consumes the Future

·
November 13, 2025
When China Consumes the Future

“There is oil in the Middle East, and there is rare earth in China”

US compute diplomacy: AI diffusion via UAE infrastructure

·
November 6, 2025
US compute diplomacy: AI diffusion via UAE infrastructure

The Situation: NVIDIA GB300 UAE Approval

User's avatar

Subscribe to Red Line

Independent intelligence on the frontier of AI and global power.

People