Frontier Safety Framework Report - Gemini 3 Pro (November, 2025) v2
storage.googleapis.com · 8,461 words · saved by 1 readers
N/A
Published: November 2025 Gemini 3 Pro Frontier Safety Framework Report Frontier Safety Framework Report – Gemini 3 Pro We released our Frontier Safety Framework (FSF) (v1) in May 2024 and updated it in February 2025 (v2) and September 2025 (v3). The FSF comprises a set of protocols that aims to address severe risks that may arise from the high-impact capabilities of frontier AI models. It currently covers four risk domains where, based on early research, we have determined severe risk may be most likely to arise from future models: CBRN (chemical, biological,…
related reading
- Claude Mythos Preview System Cardwww-cdn.anthropic.com
- Common Elements of Frontier AI Safety Policiesmetr.org
- Claude Opus 4.5: Model Card, Alignment and Safetythezvi.substack.com
- CAIS AI Dashboarddashboard.safe.ai
- A Summary of Recent Work (July 2026)gdmalignment.substack.com
- [2403.13793] Evaluating Frontier Models for Dangerous Capabilitiesarxiv.org
- Claude Mythos Preview System Cardwww-cdn.anthropic.com
- Emerging processes for frontier AI safety - GOV.UKgov.uk
- SFT Drives Gemini’s Safety Properties — LessWronglesswrong.com
- Frontier Risk Report (February to March 2026) - METRmetr.org
- GLM-5.2 Risk Evaluation Report – SaferAIsafer-ai.org
- Evaluating DeepSeek v4 Pro for Frontier Risks · Neo Researchneoresearch.ai