Skip to content

OllamaDrama: Designing and Deploying a Honeypot to Measure Attacks on Exposed LLM Infrastructure

Sep 2026 · arXiv.org · Vol abs/2609.29757 · 0 citations · 27 references
Computer Science

TL;DR

Ollure, a low- and medium-interaction honeypot that emulates the Ollama API without a backend LLM, is presented, providing empirical insight into real-world threats against exposed, self-hosted LLM services.

Abstract

Publicly exposed large language model (LLM) infrastructure creates a growing attack surface, yet real-world targeting remains poorly understood. We present Ollure, a low- and medium-interaction honeypot that emulates the Ollama API without a backend LLM. Spanning four deployments across cloud and university networks, Ollure operated for 84 days and recorded 290,887 interactions from 2,793 unique source IP addresses. Most of the activity consisted of automated discovery, fingerprinting, and model enumeration. However, we also observed concrete exploitation attempts against both the infrastructure and LLM layers. These included model management abuse, path traversal and SSRF probes, RCE and cryptocurrency mining payloads, resource exhaustion attempts, prompt injection, information extraction, and agent-oriented tool use. Our results provide empirical insight into real-world threats against exposed, self-hosted LLM services.

View source

Similar papers

Conference Open access Sep 2026

A Modular Multi-Service Honeypot for Multi-Protocol Attack Observation and Threat Intelligence Generation

Signature-based defenses have been found to be increasingly ineffective due to today’s sophisticated cyber attacks which include automation, credential guessing through brute force, web application vulnerabilities, and attacks on standard network protocols. The problem is, therefore, how attackers can be lured in order...

Yogalakshmi S, Suresh A, Ayyasamy S et al. · 0 citations
Review Open access 2026

Security of Tor Pluggable Transports: A Comprehensive Survey of Vulnerabilities, Attacks, and Countermeasures

Tor Pluggable Transports (PTs) constitute the most widely deployed censorship-circumvention layer for an estimated 2.0 to 2.5 million daily Tor users, yet their security has been assessed only through isolated, protocol-specific analyses. This survey presents the first joint vulnerability, attack, and defense framework...

Ban Alomar, Zouheir Trabelsi · 0 citations
Preprint Nov 2026

A Gingerbread House for LLM Agents: Source-Planned Honeypots That Lure Autonomous Attackers into an Isolated Shadow

LLM agents can exploit web vulnerabilities fast enough that patching may come too late. Deception against such agents has mostly lured them away from a real target that stays reachable. We study a different arrangement: an LLM defender plans a static honeypot from the application's source, and a router sends every late...

Seungdeok Roh, Seokjun Ryu, Sohee Park et al. · 0 citations
#artificial intelligence Conference Sep 2026

A2M: Trace-Optimized Agent Hijacking in the MCP Ecosystem

A2M (Attraction-to-Manipulation), a two-stage black-box framework for hijacking MCP agents, is introduced, a two-stage black-box framework for hijacking MCP agents that motivate stronger tool vetting and runtime isolation in MCP ecosystems.

Lai-Zhen Li, Xuan Wang, Pei-Cheng Zhao et al. · 0 citations
Conference Open access 2026

LLM-Powered Automated Attacks

The increasing integration of large language models (LLMs) into systems introduces new attack surfaces that extend beyond traditional software vulnerabilities. While LLMs are commonly protected by prompt-level security mechanisms, recent researches show that these controls can be bypassed through carefully crafted inpu...

Tamás Girászi, Natália Papp, Norbert Oláh et al. · 0 citations

A Large-Scale Benchmark and Risk Assessment of Traffic Analysis Attacks on Cloud LLM Services

This work presents, to its knowledge, the first unified measurement study and public benchmark of encrypted LLM traffic across both user--LLM and multi-agent executions, and assess the risk of traffic analysis attack by characterizing traffic signatures, identifying the features most associated with leakage, and testin...

Shahrooz Pouryousef, Jesus Rafael Lopez, Saeefa Rubaiyat Nowmi et al. · 2 citations · ⚡1

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.