The British AI Security Institute has reported a significant shift in the landscape of cyber capabilities, noting that open-weight models such as GLM-5.2 and DeepSeek V4-Pro have closed the performance gap with closed frontier models to just four to seven months. This is a marked improvement from the six to ten months observed at the beginning of 2025. However, the report also highlights that safety measures implemented on these open models remain largely ineffective, potentially compromising the preparedness of defenders. In this context, it is important to note that while the advancements in open-weight models present new opportunities, they also introduce risks that must be carefully evaluated. A final assessment of their long-term implications would be premature at this point, as the evolving nature of cyber threats continues to challenge organizations across various sectors.