AI Red Team Engineering notes, research references, and field notes on AI systems, responsible AI, governance, and compliance.
Global search

Search AI Red Team

Search across the complete knowledge base.

Enter at least two characters to begin.
↑↓ select Enter open Browse the full research archive
METR · October 23, 2025

Adversarial fine-tuning reviews: separate capability elicitation from risk-threshold judgments

analysis Model EvaluationAdversarial MLAI Red Teaming
Why it matters

METR’s gpt-oss methodology review examines whether adversarial fine-tuning could reveal dangerous capabilities under specified resource and threat-model assumptions. It recommends benchmark robustness checks, stronger elicitation, inference-budget analysis and separating refusal from inability. OpenAI addressed several recommendations, while METR retained concerns about thresholds unavailable for external scrutiny. The review operated under an NDA and a short implementation window; it did not assess the overall merits of releasing model weights.

My takeaway: Predefine attacker resources and decision thresholds, measure capability after reducing refusals, and document which recommendations were implemented. Keep a methodology review distinct from a complete release-risk assessment.
Keep exploring

More curated notes connected through Model Evaluation and Adversarial ML.

OpenAI News · framework

Pacing model development in an era of cyber-critical capabilities

OpenAI says preliminary evidence that Astra may meet its Critical cybersecurity threshold led it to pause frontier reinforcement-learning work for two weeks and keep its largest planned run on hold. New safeguards include stronger workload and network isolation, continuous boundary testing, token-level monitoring that escalates suspicious tool activity, and broader alignment checks for deception, reward hacking, and unauthorized access.

OWASP GenAI Security Project · guide

OWASP Top 10 for Agentic Applications for 2026

OWASP's community guide organizes agentic-system risk into ten categories, including goal hijacking, tool misuse, identity and privilege abuse, memory poisoning, insecure inter-agent communication, cascading failures, and rogue-agent behavior. It provides a shared taxonomy and mitigation starting point rather than a certification checklist or evidence that a deployed system is secure.

OpenAI News · framework

A blueprint for democratic governance of frontier AI

OpenAI proposes a three-part U.S. frontier-AI governance model: harmonize emerging state safety laws into a federal baseline, strengthen CAISI as an evaluation and standards institution, and coordinate a broader resilience program. Proposed controls include severe-risk evaluations, transparency reports, independent audits, safety-incident reporting, model-weight security, whistleblower protection, and periodic technical assessments.