Five Eyes intelligence agencies have issued a collective warning that frontier AI models capable of materially lowering the barrier for offensive cyber operations — including automated spear-phishing, AI-assisted vulnerability research, and AI-generated malware — are expected to reach public availability within months. This is a strategic planning advisory, not a discrete incident or patchable vulnerability. Note: specific model names referenced in secondary sources (Anthropic Fable 5, OpenAI Daybreak) are cited at medium confidence; the primary Five Eyes documentation was not independently verified in the source session.