Early Best Practices for Frontier AI Safety Evaluations

web

Frontier Model Forum·frontiermodelforum.org/publications/

Credibility Rating

3/5

Good(3)

Good quality. Reputable source with community review or editorial standards, but less rigorous than peer-reviewed venues.

Rating inherited from publication venue: Frontier Model Forum

The Frontier Model Forum is an industry consortium of leading AI labs; this page indexes their official publications on safety evaluation best practices, biosafety, cyber risk, and AI governance, making it a key reference for tracking industry-led safety norms.

Metadata

Importance: 62/100organizational reporthomepage

Summary

This is the publications index of the Frontier Model Forum (FMF), an industry body comprising leading AI labs, compiling their issue briefs, technical reports, research updates, and public comments on frontier AI safety. Topics span safety evaluations, biosafety thresholds, red teaming, cyber risks, compute measurement, and AI governance frameworks. It serves as a central hub for FMF's evolving best practices and policy contributions.

Key Points

•Hosts a growing library of issue briefs on frontier AI safety topics including biosafety thresholds, red teaming, chain-of-thought monitorability, and adversarial distillation.
•Technical reports cover capability assessments, frontier mitigations, third-party assessments, and cyber risk management in AI frameworks.
•Includes public comments submitted to US government bodies (NIST, CAISI, AISI) on AI safety standards and policy.
•Represents collaborative output from major frontier AI developers (Anthropic, Google, Microsoft, OpenAI) on shared safety norms.
•Publications date from 2023 onward, reflecting the rapid evolution of industry-led AI safety evaluation practices.

Cited by 2 pages

Page	Type	Quality
Frontier AI Labs (Overview)	--	85.0
AI Lab Safety Culture	Approach	62.0

Cached Content Preview

HTTP 200Fetched Apr 7, 20263 KB

Publications - Frontier Model Forum

Publications

Issue Briefs

Adversarial Distillation. February 23, 2026.

Chain of Thought Monitorability . January 27, 2026.

Preliminary Taxonomy of AI-Bio Misuse Mitigations . July 30, 2025.

Frontier AI Biosafety Thresholds . May 12, 2025.

Preliminary Reporting Tiers for AI-Bio Safety Evaluations. March 18, 2025.

Thresholds for Frontier AI Safety Frameworks . February 7, 2025.

Preliminary Taxonomy of AI-Bio Safety Evaluations . December 20, 2024.

Preliminary Taxonomy of Pre-Deployment Frontier AI Safety Evaluations . December 20, 2024.

AI for Cyber Defense . November 22, 2024.

Components of Safety Frameworks . November 8, 2024.

Early Best Practices for Frontier AI Safety Evaluations . July 31, 2024.

Foundational Security Practices . July 31, 2024.

Measuring Training Compute . May 2, 2024.

What is Red Teaming? . October 24, 2023.

Technical Reports

Managing Advanced Cyber Risks in Frontier AI Frameworks . February 13, 2026.

Third-Party Assessments . August 4, 2025.

Frontier Mitigations . June 30, 2025.

Risk Taxonomy and Thresholds for Frontier AI Frameworks . June 18, 2025.

Frontier Capability Assessments . April 22, 2025.

Research Updates

Frontier AI and Nuclear Security . January 30, 2026.

Public Comments

FMF Response to the US CAISI Request for Information Regarding Security Considerations of AI Agents . March 9, 2026.

FMF Comment on NIST Outline of Draft TEVV Standard . September 12, 2025.

FMF Response to the Request for Information on the Development of an AI Action Plan . March 14, 2025.

FMF Comment on US AISI RFI: Safety Considerations for Chemical and/or Biological AI Models . December 3, 2024.

FMF Response to Managing Misuse Risk for Dual-Use Models, NIST Report 800-1 . September 9, 2024.