Open-weight models now match frontier cyber performance from just four months ago at a fraction of the cost

The British AI Security Institute reports that open-weight models like GLM-5.2 and DeepSeek V4-Pro are closing the cyber capability gap with closed-source frontier models, trailing them by only four to seven months compared to the six-to-ten-month lag seen early this year. This progress indicates that high-level offensive capabilities are becoming significantly more accessible and affordable. However, the study also concludes that safety guardrails on these open models remain largely ineffective at preventing misuse.