Our Mission

MidnightAI.org is an independent research initiative dedicated to monitoring and analyzing the global progress toward artificial general intelligence (AGI) and superintelligent AI. Inspired by the Bulletin of the Atomic Scientists' Doomsday Clock, we provide a data-driven assessment of how close humanity is to achieving transformative AI capabilities.

What We Do

Real-Time Monitoring

Continuous analysis of AI developments, model releases, and capability breakthroughs from major research labs worldwide.

AI-Powered Analysis

Leveraging advanced AI to synthesize news, research papers, and announcements into actionable intelligence assessments.

Capability Tracking

Systematic assessment of AI capabilities across key domains: reasoning, coding, agency, multimodal understanding, and more.

Open Research

Transparent methodology and data sharing to support the broader AI safety and governance research community.

The Doomsday Clock

Our “Minutes to Midnight” metric represents an estimated timeline to the emergence of superintelligent AI. This is calculated using a weighted analysis of:

  • Current AI capability levels across seven key domains
  • Rate of progress and acceleration patterns
  • Recent breakthroughs and their significance
  • Investment and resource allocation trends
  • Expert assessments and research publications

The clock is updated regularly as new developments emerge, providing a dynamic view of the AI landscape.

Organizations We Monitor

We track developments from leading AI research organizations worldwide, including:

United States

OpenAIAnthropicGoogle DeepMindMeta AIxAIMicrosoft AI

China

DeepSeekAlibaba (Qwen)ByteDanceBaidu (Ernie)

Europe

Mistral AIAleph Alpha

Canada & Israel

CohereAI21 Labs

Support Our Mission

MidnightAI.org is an independent, community-funded research initiative. Your support helps us maintain our commitment to transparent, unbiased tracking of AI progress.

Disclaimer

MidnightAI.org is an independent research project for educational and informational purposes. Our assessments are based on publicly available information and should not be considered definitive predictions. The timeline to superintelligent AI remains highly uncertain, and our analyses reflect current understanding subject to revision as new information becomes available.