ai-safety
Safeguard articles tagged "ai-safety" — guides, analysis, and best practices for software supply chain and application security.
6 articles
Cisco's May 2026 Multi-Turn Jailbreak Study: Why Frontier Model Safety Collapses Over a Conversation
Cisco's AI threat team tested 15 flagship models with ~7,000 multi-turn attacks and found success rates as high as 88 percent. Single-turn safety scores told defenders almost nothing about real-world resilience.
LLM Jailbreak Defense Architectures in 2026
Jailbreaks against frontier models keep getting more sophisticated. The defense architectures that have proven durable, and the ones that get bypassed in weeks.
AI Safety Eval Datasets as Supply Chain
The datasets you use to evaluate model safety are themselves a supply chain, and almost nobody is treating them that way. A senior engineer's audit of how eval corpora get poisoned, contaminated, and silently drifted.
Gemini 2.5 Pro and the Late Safety Report
Google released Gemini 2.5 Pro Experimental on March 25, 2025 without a contemporaneous safety report. The UK reaction set a precedent.
GPAI Code of Practice: The 2025 Signatory Landscape
The General-Purpose AI Code of Practice was published on 10 July 2025 with three chapters. Most major providers signed, with notable partial signatures from xAI.
OpenAI Preparedness Framework v2: April 2025 Update
OpenAI released Preparedness Framework v2 on April 15, 2025 with sharper thresholds, an AI self-improvement category, and clearer disclosure requirements. We unpack the operational changes.