Autonomy Risk Countdown

This site is a public monitor for autonomous systems risk. It estimates when current trends in AI capability, military integration, human oversight, governance, and real-world incidents could converge into a high-risk autonomy threshold where meaningful human control begins to break down.

What This Site Tracks

• Frontier AI capability and multi-agent autonomy
• Military and institutional integration of autonomous systems
• Human override capacity and governance enforcement gaps
• Current incidents, source-backed evaluations, and estimated risk horizons

What Visitors Can See

• A live countdown to the model's current estimated risk horizon
• A recent incident feed with source links
• A history of past evaluations and how the estimate has shifted over time
• Underlying model notes, indicators, and projection details

The interactive version of the monitor is available at https://ctrl-c.nl/autonomy-risk/.

A crawler-friendly summary is also available at summary.html and summary.txt.