# Tiny Enough to Break In: Agentic Remote Access Trojans Powered by Small Language Models

arXiv: https://arxiv.org/abs/2608.03009 · HTML: https://mlsanity.com/paper/2608.03009/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Agent threats, guardrails and governance, one of 16 directions in LLM agents and tool use.
- Builds on 18 of its 33 references that are in the map.
- First posted on arXiv 4 Aug 2026.

## Placed in

- LLM agents and tool use / [Agent threats, guardrails and governance](https://mlsanity.com/agents/agent-threats-guardrails-and-governance/)

## Builds on (most cited first, 18 of 18)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [Understanding the planning of LLM agents: A survey](https://mlsanity.com/paper/2402.02716/) (2024)
- [Small Language Models are the Future of Agentic AI](https://mlsanity.com/paper/2506.02153/) (2025)
- [Cybench: A Framework for Evaluating Cybersecurity Capabilities and Risks of Language Models](https://mlsanity.com/paper/2408.08926/) (2024)
- [LLM Agents can Autonomously Exploit One-day Vulnerabilities](https://mlsanity.com/paper/2404.08144/) (2024)
- [LLM Agents can Autonomously Hack Websites](https://mlsanity.com/paper/2402.06664/) (2024)
- [PentestGPT: An LLM-empowered Automatic Penetration Testing Tool](https://mlsanity.com/paper/2308.06782/) (2023)
- [AutoAttacker: A Large Language Model Guided System to Implement Automatic Cyber-attacks](https://mlsanity.com/paper/2403.01038/) (2024)
- [Getting pwn'd by AI: Penetration Testing with Large Language Models](https://mlsanity.com/paper/2308.00121/) (2023)
- [Hackphyr: A Local Fine-Tuned LLM Agent for Network Security Environments](https://mlsanity.com/paper/2409.11276/) (2024)
- [SlimLM: An Efficient Small Language Model for On-Device Document Assistance](https://mlsanity.com/paper/2411.09944/) (2024)
- [Architecting Secure AI Agents: Perspectives on System-Level Defenses Against Indirect Prompt Injection Attacks](https://mlsanity.com/paper/2603.30016/) (2026)
- [Learning Explainable Representations of Malware Behavior](https://mlsanity.com/paper/2106.12328/) (2021)
- [A Survey of LLM-Driven Penetration Testing: Taxonomy, Co-Evolution, and Open Challenges](https://mlsanity.com/paper/2607.02605/) (2026)
- [Exploit Kits: The production line of the Cybercrime Economy](https://mlsanity.com/paper/1707.04714/) (2017)
- [Ransomware 3.0: Self-Composing and LLM-Orchestrated](https://mlsanity.com/paper/2508.20444/) (2025)

## Built on by (most cited first, 1 of 1)

- [Big Enough to Break Out: Tracking the Rising Capability of LLM Penetration-Testing Agents](https://mlsanity.com/paper/2609.10780/) (2026)

## Abstract (arXiv)

Agentic artificial intelligence raises a new security concern: cyber threats that reason, act, and adapt locally without continuous human direction. We examine this threat through an Agentic Remote Access Trojan (agentic RAT): a Remote Access Trojan augmented with a locally deployed Small Language Model (SLM). The SLM interprets host and network observations, selects actions, recovers from failed steps, and reduces reliance on an external operator. We implement the concept in a controlled, network-isolated lab built from Kali Linux, a Metasploitable2 target, LM Studio, and a local 8-billion-parameter Dolphin-family model. We then test whether a model this small can support autonomous cyber decision-making. This is architecturally feasible today. On commodity hardware, with no cloud service and no operator in the loop, the SLM closed the full observe-decide-act cycle: it interpreted ranked reconnaissance evidence supplied by the controller, selected actions, and obtained verified root-shell access on real vulnerable services. However, it is not yet operationally reliable. The same model hallucinated commands, misread output, and recovered from failure inconsistently, completing 10.9% of a deliberately strict checklist. That gap reflects the limits of today's small models, not a ceiling on the concept. As SLMs improve, agentic endpoint systems may become more practical, more autonomous, and harder to detect, straining existing monitoring, containment, and policy-enforcement mechanisms. Real-world incidents in 2025-2026 already show AI-driven intrusions moving from concept toward practice. That makes the local, self-contained variant we study a plausible near-term direction, not a hypothetical one.