Safety · Trend
Narrowing Cyber Capability Gap of Open-Weight Models
AISI finds leading open-weight models trail frontier closed cyber capability by only 4–7 months, down from 6–10.
Forecast horizon
Open-weight cyber parity may compress defender lead times for evaluation and export controls.
Connections
Connections · 3
How this node ties into the rest of the map, and the evidence behind each link.
from · tracked by 4/10
UK AI Safety Institute AISI cyber evaluation finds open-weight models only 4–7 months behind frontier closed models.
+4 growth from · informs 4/10
ORCA-bench: Oncall Root-Cause Analysis for Coding Agents Low RCA accuracy of frontier coding agents under realistic oncall noise complements cyber-capability gap measurements for operational risk.
+4 growth to · references 3/10
DeepSeek-V4: Million-Token MoE Language Model AISI cites DeepSeek V4-Pro among open models performing near frontier closed cyber capability.
+3 growthSignal sources
Signal sources
Dated facts from primary sources in this direction.
US evaluation centre Jun 2025
In June 2025 the US AI Safety Institute was renamed the Center for AI Standards and Innovation (CAISI), pivoting toward security, standards and adversary-model assessment.
NIST → Frontier safeguards May 2025
Anthropic activated its ASL-3 deployment and security standard with Claude Opus 4 on 22 May 2025 — the first real-world trigger of a responsible-scaling tier, focused on blocking bio-weapon uplift.
Anthropic → Cross-border testing 2025
The International Network of AI Safety Institutes (launched Nov 2024) ran a third joint testing exercise focused on agentic AI systems across cyber and fraud strands.
European Commission — AI Office →