This frontier framework is threshold-centric and public on named severe-risk domains, but it provides no self-harm category or self-harm-specific governance machinery. Publication date is coded at month precision because the source states only Published: November 2025.
Contents
Summary
Frontier Safety Framework Report - Gemini 3 Pro (November, 2025) v2 is a frontier safety framework from Google DeepMind dated November 2025. This frontier framework is threshold-centric and public on named severe-risk domains, but it provides no self-harm category or self-harm-specific governance machinery. Publication date is coded at month precision because the source states only Published: November 2025.
Representation
This document codes self-harm as absent with evaluative-object status Absent / not represented. The public crosswalk records self-harm naming as absent and mechanism coverage as none.
| Field | Value |
|---|---|
| Organization | Google DeepMind |
| Document type | frontier_safety_framework |
| Document genre | frontier_or_scaling |
| Risk status | Absent / not represented |
| First-class equivalence | no |
| Threshold mechanism | none |
| Deployment consequence | none |
| Multi-turn evaluation | none |
Assessment
The retained crosswalk places this document at Absent / not represented. That classification reflects what the document itself operationalizes publicly, not the organization's overall internal practice.
Related pages
Use the linked provider, methods, and neighboring governance pages below to place this document in context.
- Google and Google DeepMind
- Coding Rules
- Source Selection
- ANTHROPIC-2025-04
- ANTHROPIC-2026-01
- GOOGLEDEEPMIND-2025-02
Sources
The primary public source for this page is https://deepmind.google/models/fsf-reports/gemini-3-pro/. The coded evidence excerpt is: It currently covers four risk domains ... CBRN ..., cybersecurity, machine learning R&D, and harmful manipulation, and also includes ... misalignment risk..
This page is grounded in the DOI policy registry, the policy coding manual, and the document's own public URL.
Access publication
Safety note
This publication preserves a harm-minimization posture and does not provide self-harm instructions. If you or someone you know is in immediate danger, call your local emergency number. In the U.S. and Canada, call or text 988.