What OpenAI Preparedness Framework 2023 - Model Capability Evaluation: Cybersecurity, CBRN, Persuasion and Model Autonomy Risk Scoring, Critical (Red)/High/Medium/Low Thresholds, Safety Advisory Group Veto Power, Post-Mitigation Score Requirement Before Deployment requires
This framework requires developers of advanced AI models to evaluate and score capabilities in high-risk domains including cybersecurity, chemical-biological-radiological-nuclear (CBRN), persuasion, and model autonomy using defined risk thresholds. Deployment is prohibited if post-mitigation risk scores remain at Critical (Red) level, subject to Safety Advisory Group veto authority.
Pillar: AI Governance & Law · Authority: OpenAI · Version: 1.0.1 · Last updated:
Primary source: https://openai.com/safety/preparedness
SHA-256 integrity: a53cd52046533bbc008fbb580af67cab6c9ad610226f5f4dda88f39ae7a09607
Primary Citations — 5 traced to source
- OpenAI Preparedness Framework 2023, https://openai.com/safety/preparedness, see source document, 2023
- OpenAI Preparedness Framework 2023, https://openai.com/safety/preparedness, see source document, 2023
+ 3 more citations (full bibliography, deterministic workflow, actionable schema and crosswalks) included in the vault unlock — $0.01 via Skyfire / L402 / Direct Base USDC.
Access
- Discovery (free): /api/v1/nodes/openai-preparedness-framework-2023.json — 6-field metadata
- Vault (full node): /api/v1/vault/nodes/openai-preparedness-framework-2023.json — full 13-key payload, $0.01 USDC (L402/Skyfire/Direct Base)
- Canonical URL: https://bidda.com/intelligence/openai-preparedness-framework-2023
- Back to registry: Browse all 10,099 compliance nodes