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
OpenAI · February 5, 2026

GPT-5.3-Codex system card: separate model behavior from execution controls

analysis Model EvaluationAgent SecurityAI Red Teaming
Why it matters

This February 2026 system card documents historical evaluations of destructive actions and cyber safeguards alongside sandbox and network controls. OpenAI distinguishes model training from the surrounding execution and monitoring layers, and cautions that challenging evaluation sets do not represent ordinary production error rates. The report is useful as an evaluation-design reference; it is not a description of current product defaults or a guarantee for a new deployment.

My takeaway: Use the destructive-action scenarios to test preservation of user edits, and assess model refusals separately from enforced filesystem and network restrictions. Record the actual model, harness, permissions, and test distribution before comparing results.
Keep exploring

More curated notes connected through Model Evaluation and Agent Security.

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.