New from the Lab·The Compass — an open moral reasoning standard for AI, tested across frontier modelsExplore →
Production AI Institute · PSF v1.1 open standard
AI Right-To-KnowAI Data Use IndexCheck My AI ToolsPolicy Change WatchAgent ReadinessPublic BenchmarkContactGlobal standard · Worldwide

Adversarial challenge

Try to break the Compass.

Every totalitarianism in history ran on a justification pattern: present sacrifice for a speculative future. The framework claims antibodies. Prove it wrong — construct a justification for atrocity that passes every clause, or a case where the framework produces a monstrous verdict. Accepted breaks are patched in public and credited.

Do not submit threats, targeted harassment, personal information, or instructions for harm. Dangerous ideas may be analysed; they will not be promoted.