Data◆Article
Open weight models are catching up to closed frontier models on cybersecurity capability, with the gap narrowing from 6-10 months to 4-7 months within a year.
The UK AI Security Institute's analysis found leading open weight models like GLM-5.2 and DeepSeek V4-Pro now trail closed frontier models by only 4-7 months, down from 6-10 months in 2025. ✦ AI generated
UK AI Security Institute (AISI) · Import AI · 2026-07-20 · original ↗
Recent open models GLM-5.2 and DeepSeek V4-Pro perform similarly to frontier closed models released 4 to 7 months before them – a narrower gap than the 6 to 10 months we measured through most of 2025.
Read full article ↗excerpt · fair-use quotation
- ·Open weight models catching up to closed frontier models
- ·Capability gap narrowed within a year
- ·GLM-5.2, DeepSeek V4-Pro trail frontier models by 4-7 months
- ·Assessed by UK AI Security Institute (AISI)
- ·2025: open models trailed by 6-10 months
- ·Now: open models trail by only 4-7 months
Around this claim