---
title: "LLM-Guided Cyber-Physical Attack Path... | SI Paper Desk"
url: https://thesimind.si/papers/2611.24765/
description: "The authors use language models with structured threat modeling to find likely attack paths in power grids, and report closer agreement with physics-based..."
---

Number 12 of 12 · November 20, 2026

# LLM-Guided Cyber-Physical Attack Path Identification in Smart Grids

The authors use language models with structured threat modeling to find likely attack paths in power grids, and report closer agreement with physics-based rankings and much faster runtime than a fusion baseline.

Amr Elshazly, Jesse Roberts, Rachad Atat and 1 more · Submitted November 18, 2026 · arXiv 2611.24765v1 · cs.CR

[Read the paper on arXiv](https://arxiv.org/abs/2611.24765) · [The PDF](https://arxiv.org/pdf/2611.24765) · [The day's 12](https://thesimind.si/papers/day/2026-11-20/)

## What they found

The authors built a framework that uses large language models (LLMs) to find attack paths in a power grid's cyber and physical systems. It combines grid topology (how parts connect) with LLM-guided STRIDE and MITRE ATT&CK threat modeling. They compared it with three baselines: topology-driven, threat-driven, and fusion-based. They checked results against rankings taken from cascading-failure simulations, which show how a fault spreads. In the cyber layer, critical-node overlap@3 rose from approximately 33% to as high as 100%. Node-level similarity reached up to 65% and path-level alignment up to 48%. Runtime fell by approximately 75-85% against the fusion-based baseline in the IEEE 14-bus system.

## Why it matters

Power grid operators need to know which attacks would do the most damage. Topology tools miss how an attacker thinks. Simulations are slow when threats keep changing. This work suggests an LLM could give faster, impact-aware answers. The abstract reports tests on one small system, the IEEE 14-bus system, so how well it scales is not stated. Path-level alignment of up to 48% also means many full attack sequences still differ from the reference.

## How it changes the race

This is a niche application of SI to security analysis, not a change in frontier capability. If it holds, it shows that SI models can give faster first-pass risk screening for critical infrastructure.

## Safety and control

The work is defensive and aims to help find high-impact attack paths so grids can be made more resilient. The same kind of attack path ranking could also help an attacker, though the abstract does not discuss this.

### The abstract, by the authors

Identifying high-impact attack paths in cyber-physical power systems is critical for improving grid resilience. Existing methods typically rely on either structural topology analysis or simulation-based cascading-failure models. Topology-based approaches capture connectivity but lack adversarial reasoning, while simulation-driven methods are primarily suited for offline analysis and are less effective for rapid evaluation of evolving attack scenarios. As a result, identifying realistic attack paths that jointly capture topology, threat semantics, and physical impact remains challenging. This paper presents a physics-grounded framework for identifying cyber-physical attack paths using large language models (LLMs), with cascading-failure-derived impact rankings used as physics-grounded references. The framework integrates system topology, LLM-guided STRIDE and MITRE ATT&CK threat modeling, and three baseline strategies for comparison: topology-driven, threat-driven, and fusion-based. Evaluation is performed using (i) critical-node overlap between LLM-inferred top-K rankings and cascading-failure-based ground truth; (ii) node-level similarity between LLM-inferred and reference paths; and (iii) path-level alignment capturing agreement in complete attack sequences. Results show that structured LLM-based threat modeling substantially improves alignment with physics-grounded references. In the cyber layer, critical-node overlap@3 increases from approximately 33% to as high as 100%, while node-level similarity reaches up to 65% and path-level alignment reaches up to 48%. In addition, the proposed framework reduces runtime by approximately 75-85% compared to the fusion-based baseline in the IEEE 14-bus system. These results suggest that LLM-guided attack path identification is a promising approach for efficient and impact-aware cyber-physical security analysis.

Quoted from arXiv:2611.24765. The research and its words belong to its authors.

Desk score **73** out of 100

Why it was picked

- Alignment
- Security (in the title)
- Reasoning
- Evals
- Language models (in the title)
- Cross-listed in 2 fields

Topics

- [Security](https://thesimind.si/papers/?tag=security)
- [Language models](https://thesimind.si/papers/?tag=language-models)
- [Evals](https://thesimind.si/papers/?tag=evals)
- [Agents](https://thesimind.si/papers/?tag=agents)
- [Science](https://thesimind.si/papers/?tag=science)

**arXiv ID:** 2611.24765v1
**Submitted:** November 18, 2026
**Fields:** cs.CR (Cryptography and Security), cs.AI (Artificial Intelligence)
**Level:** Intermediate
**Comments:** 7 pages, 4 figures, 1 table. Accepted at IEEE IECON 2026

Fingerprint, SHA-256

`33cd95c5af2d9e2bb43d61e7737475ade9d1f99baa6b9a51b2f0292e187cf8dd`

On Solana: [Proof of Mind seal, 4JtzxB...6w5n](https://solscan.io/tx/4JtzxBBunUdHpdT7ydTygvUny81ph9bmXUMt9UHMJBLFrSwBZFdTbtV5CcZdHAM2kg7Bc2pTnq59P4DvXyGc6w5n)

### The exact text fingerprinted

{"arxiv":"2611.24765","version":1,"title":"LLM-Guided Cyber-Physical Attack Path Identification in Smart Grids","authors":["Amr Elshazly","Jesse Roberts","Rachad Atat","Muhammad Ismail"],"day":"2026-11-20","rank":12,"score":73,"status":"written","summary":{"takeaway":"The authors use language models with structured threat modeling to find likely attack paths in power grids, and report closer agreement with physics-based rankings and much faster runtime than a fusion baseline.","found":"The authors built a framework that uses large language models (LLMs) to find attack paths in a power grid's cyber and physical systems. It combines grid topology (how parts connect) with LLM-guided STRIDE and MITRE ATT&CK threat modeling. They compared it with three baselines: topology-driven, threat-driven, and fusion-based. They checked results against rankings taken from cascading-failure simulations, which show how a fault spreads. In the cyber layer, critical-node overlap@3 rose from approximately 33% to as high as 100%. Node-level similarity reached up to 65% and path-level alignment up to 48%. Runtime fell by approximately 75-85% against the fusion-based baseline in the IEEE 14-bus system.","matters":"Power grid operators need to know which attacks would do the most damage. Topology tools miss how an attacker thinks. Simulations are slow when threats keep changing. This work suggests an LLM could give faster, impact-aware answers. The abstract reports tests on one small system, the IEEE 14-bus system, so how well it scales is not stated. Path-level alignment of up to 48% also means many full attack sequences still differ from the reference.","race":"This is a niche application of SI to security analysis, not a change in frontier capability. If it holds, it shows that SI models can give faster first-pass risk screening for critical infrastructure.","safety":"The work is defensive and aims to help find high-impact attack paths so grids can be made more resilient. The same kind of attack path ranking could also help an attacker, though the abstract does not discuss this.","level":"intermediate","tags":["security","language-models","evals","agents","science"]}}

[Check it on Verify a thought](https://thesimind.si/proof-of-mind/)

## The authors

Amr Elshazly, Jesse Roberts, Rachad Atat, Muhammad Ismail

Credit for this research belongs to its authors. SI MIND's page is a plain-English reading of their abstract on arXiv, not a review of the full paper.

Related on the desk

## More papers on these topics

[Nov 16, 2026 · Paper of the DayOnly as Safe as the Weakest Agent: A Break That Appears Only When the System Is Multi-AgentWhen a strong SI model hands work to a cheaper sub-agent, the whole system can be only as safe as the weaker agent, even if the strong model alone resists the attack.SafetyMulti-agent systemsSecurity](https://thesimind.si/papers/2611.17520/) · [Nov 20, 2026 · Number 2AutoPI: Taxonomy-Guided Adaptive Red Teaming for Prompt Injection in LLM AgentsAutoPI is a red teaming tool that tests SI agents for prompt injection across many attack types, not just how often attacks work, and it reports higher success and coverage than recent rivals.SecurityAgentsEvals](https://thesimind.si/papers/2611.24313/) · [Nov 16, 2026 · Number 2Soteria: Formally Verified Planning with Runtime Enforcement for Safe LLM AgentsSoteria checks an SI agent's plan against written rules before any tool runs, then enforces those rules during execution. The authors report more completed tasks and no policy violations.AgentsSafetyReasoning](https://thesimind.si/papers/2611.17880/)

[Number 11](https://thesimind.si/papers/2611.23403/) · [All 12 for Nov 20, 2026](https://thesimind.si/papers/day/2026-11-20/)
