Import AI 465: Open vs closed gaps; Kimi K3; Demis' big policy plan
Open-weight models now trail closed frontier models by only 4-7 months on cybersecurity capabilities
“This implies cyber defenders have a short window to prepare before today's frontier cyber capabilities may become accessible without the same safeguards”— Jack Clark
The UK AI Security Institute has documented that the gap between open-weight and closed frontier models on cybersecurity tasks has shrunk from 6-10 months to just 4-7 months in 2025, with GLM-5.2 and DeepSeek V4-Pro approaching recent Anthropic and OpenAI releases. Separately, China's Kimi K3 (2.8T parameters) now matches or narrowly trails Claude Fable 5 and GPT 5.6 Sol on major benchmarks. The combined signal is that frontier-level cyber capabilities — currently controlled by companies with safety guardrails — are rapidly diffusing into the open ecosystem, narrowing the window for defenders to prepare.