Novel "Policy Puppetry Attack" Bypasses Safety Guardrails in Major AI Models, Raising Urgent Security Concerns

Novel Universal Bypass for All Major LLMs

Summary Researchers at HiddenLayer have developed the first, post-instruction hierarchy, universal, and transferable prompt injection technique that successfully bypasses instruction hierarchy and safety guardrails across all major frontier AI models. This includes models from OpenAI (ChatGPT 4o, 4o-mini, 4.1, 4.5, o3-mini, and o1), Google (Gemini 1.5, 2.0, and 2.5), Microsoft (Copilot), Anthropic (Claude 3.5 and 3.7), Meta (Llama 3 and 4 families), DeepSeek (V3 and R1), Qwen (2.5 72B) and Mistr...