Safety · Concept
Fundamental Flaw Leaving LLMs Attack-Vulnerable
Report that LLMs cannot be made fully secure against hacks because of a structural flaw in how they operate.
Forecast horizon
Strengthens case for layered external controls over model-only hardening.
Connections
Connections · 2
How this node ties into the rest of the map, and the evidence behind each link.
to · drives 4/10
Industry-Wide Jailbreak Severity Scoring Framework If models are architecturally un-hardenable, industry severity scoring and external controls become more central.
+4 growth to · enables 4/10
CoT-Hidden Backdoors via Model Poisoning Structural LLM attack surface claims align with demonstrations that poisoned models can hide malicious intent in benign CoT.
+4 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 →