Model Releases

AISI's First Public Evaluation: Leading Open-Weight Models Lag Frontier by 4 to 7 Months in Cybersecurity Capabilities

Original title:How Far Behind the Frontier are Leading Open Weight Models on Cyber?

AI summary

The UK AI Security Institute published its first public analysis of the cybersecurity capability gap of leading open-weight models. Testing found that GLM-5.2 was the strongest open-weight model at test time, approaching Opus 4.6 released four months earlier on AISI narrow-scope cyber tasks and approaching Opus 4.5 on long-horizon cyber ranges, overall lagging the frontier by 4 to 7 months, narrower than the 6 to 10 months measured in internal evaluations from January to September 2025.

Key Takeaways

  • The UK AI Security Institute published its first public analysis of the cybersecurity capability…
  • Testing found that GLM-5.
  • 2 was the strongest open-weight model at test time,approaching Opus 4.
  • 6 released four months earlier on AISI narrow-scope cyber tasks and approaching Opus 4.
  • 5 on long-horizon cyber ranges,overall lagging the frontier by 4 to 7 months,narrower than the 6…

Quick Facts

Published
Updated
Source
英国 AI Security Institute:Blog(网页)
Category
Model Releases
AI score
70
Read the original Read the full article at the original source
Why it matters

这是 AISI 首次公开量化开源权重模型与闭源前沿在网络安全能力上的差距,并给出成本对比,为防御方的准备窗口提供依据。

Related entities

Source & copyright

This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.

Original article · AIHOT entry

How to Cite

When citing this article, keep the title, date, and permalink, and credit Orange News as the source.

Orange News: AISI's First Public Evaluation: Leading Open-Weight Models Lag Frontier by 4 to 7 Months in Cybersecurity Capabilities. July 17, 2026. https://new.nat6.net/en/news/ga76bea2cgzi87tercdip10qu/

View all →