The Global Open Competition
The
AI Alignment
challenge
Critiqued and Judged by four frontier AI models (Claude Opus 4.6, ChatGPT 5.4, Gemini 3.1, Grok 4.2). Scored on a transparent 100-point rubric: 50 points for Core Merit (soundness, novelty, effectiveness, failure analysis, honesty), 50 points for the Alignment Razor (breadth and elegance). A great AI alignment submission must address everything — including the hard cases.
Submit Your Idea
1
Sperman Alignment Framework v1.1James Sperman✦ scored
Core Merit28/50
Razor42/50
70
/100
▼
2
Ontological Containment Framework (OCF) v1.0Gemini 3.1 Pro✦ scored
Core Merit27/50
Razor39/50
65
/100
▼
3
ChatGPT's Alignment Plan v1.0ChatGPT 5.4✦ scored
Core Merit22/50
Razor42/50
64
/100
▼
4
Agent Foundations (MIRI) v1.0Soares, Fallenstein, et al✦ scored
Core Merit22/50
Razor38/50
60
/100
▼
5
AI Safety Via Debate v1.0Irving and Christiano✦ scored
Core Merit19/50
Razor39/50
58
/100
▼
6
Eliciting Latent Knowledge (ELK) v1.0Evan Hubinger, et al✦ scored
Core Merit17/50
Razor40/50
57
/100
▼
7
Truth-Seeking AI Alignment v1.0Grok 4.2✦ scored
Core Merit19/50
Razor38/50
57
/100
▼
8
Claude's Alignment Plan v1.0Claude Opus 4.6✦ scored
Core Merit16/50
Razor35/50
51
/100
▼
9
Alignment-Aware-Neural-Architecture--AANA-Armando Sori✦ scored
Core Merit17/50
Razor33/50
50
/100
▼
10
Cooperative AI Framework v1.0Dafoe et al✦ scored
Core Merit17/50
Razor33/50
50
/100
▼
11
Iterated Amplification v1.0Paul Christiano✦ scored
Core Merit12/50
Razor32/50
44
/100
▼
Think your idea can compete?
Submit your AI alignment idea for AI-powered evaluation on the same transparent rubric. Every entry receives a detailed constructive critique. $20 entry fee.
Scored by Claude Opus 4.6 (Anthropic), ChatGPT 5.4 (OpenAI), Gemini 3.1 Pro (Google), and Grok 4.2 (xAI) — averaged for fairness.
All scoring justifications are transparent and visible. Click any entry to see details.
© 2026 AI Alignment Challenge