Search

Large Language Models (LLMs) have shown remarkable potential across various domains, including cybersecurity. Using commercial cloud-based LLMs may be undesirable due to privacy concerns, costs, and network connectivity constraints. In this paper, we present Hackphyr, a locally fine-tuned LLM to be used as a red-team agent within network security environments. Our fine-tuned 7 billion parameter model can run on a single GPU card and achieves performance comparable with much larger and more powerful commercial models such as GPT-4. Hackphyr clearly outperforms other models, including GPT-3.5-turbo, and baselines, such as Q-learning agents in complex, previously unseen scenarios. To achieve this performance, we generated a new task-specific cybersecurity dataset to enhance the base model's capabilities. Finally, we conducted a comprehensive analysis of the agents' behaviors that provides insights into the planning abilities and potential shortcomings of such agents, contributing to the broader understanding of LLM-based agents in cybersecurity contexts

Honeypots are essential tools in cybersecurity for early detection, threat intelligence gathering, and analysis of attacker's behavior. However, most of them lack the required realism to engage and fool human attackers long-term. Being easy to distinguish honeypots strongly hinders their effectiveness. This can happen because they are too deterministic, lack adaptability, or lack deepness. This work introduces shelLM, a dynamic and realistic software honeypot based on Large Language Models that generates Linux-like shell output. We designed and implemented shelLM using cloud-based LLMs. We evaluated if shelLM can generate output as expected from a real Linux shell. The evaluation was done by asking cybersecurity researchers to use the honeypot and give feedback if each answer from the honeypot was the expected one from a Linux shell. Results indicate that shelLM can create credible and dynamic answers capable of addressing the limitations of current honeypots. ShelLM reached a TNR of 0.90, convincing humans it was consistent with a real Linux shell. The source code and prompts for replicating the experiments have been publicly available., Comment: 6 pages. 2 figures. 2 tables

Due to the proliferation of malware, defenders are increasingly turning to automation and machine learning as part of the malware detection tool-chain. However, machine learning models are susceptible to adversarial attacks, requiring the testing of model and product robustness. Meanwhile, attackers also seek to automate malware generation and evasion of antivirus systems, and defenders try to gain insight into their methods. This work proposes a new algorithm that combines Malware Evasion and Model Extraction (MEME) attacks. MEME uses model-based reinforcement learning to adversarially modify Windows executable binary samples while simultaneously training a surrogate model with a high agreement with the target model to evade. To evaluate this method, we compare it with two state-of-the-art attacks in adversarial malware creation, using three well-known published models and one antivirus product as targets. Results show that MEME outperforms the state-of-the-art methods in terms of evasion capabilities in almost all cases, producing evasive malware with an evasion rate in the range of 32-73%. It also produces surrogate models with a prediction label agreement with the respective target models between 97-99%. The surrogate could be used to fine-tune and improve the evasion rate in the future., Comment: 12 pages, 3 figures, 3 tables. Accepted at ESORICS 2023

Ransomware-as-a-service (RaaS) is increasing the scale and complexity of ransomware attacks. Understanding the internal operations behind RaaS has been a challenge due to the illegality of such activities. The recent chat leak of the Conti RaaS operator, one of the most infamous ransomware operators on the international scene, offers a key opportunity to better understand the inner workings of such organizations. This paper analyzes the main topic discussions in the Conti chat leak using machine learning techniques such as Natural Language Processing (NLP) and Latent Dirichlet Allocation (LDA), as well as visualization strategies. Five discussion topics are found: 1) Business, 2) Technical, 3) Internal tasking/Management, 4) Malware, and 5) Customer Service/Problem Solving. Moreover, the distribution of topics among Conti members shows that only 4% of individuals have specialized discussions while almost all individuals (96%) are all-rounders, meaning that their discussions revolve around the five topics. The results also indicate that a significant proportion of Conti discussions are non-tech related. This study thus highlights that running such large RaaS operations requires a workforce skilled beyond technical abilities, with individuals involved in various tasks, from management to customer service or problem solving. The discussion topics also show that the organization behind the Conti RaaS oper5086933ator shares similarities with a large firm. We conclude that, although RaaS represents an example of specialization in the cybercrime industry, only a few members are specialized in one topic, while the rest runs and coordinates the RaaS operation.

Large Language Models (LLMs) have gained widespread popularity across diverse domains involving text generation, summarization, and various natural language processing tasks. Despite their inherent limitations, LLM-based designs have shown promising capabilities in planning and navigating open-world scenarios. This paper introduces a novel application of pre-trained LLMs as agents within cybersecurity network environments, focusing on their utility for sequential decision-making processes. We present an approach wherein pre-trained LLMs are leveraged as attacking agents in two reinforcement learning environments. Our proposed agents demonstrate similar or better performance against state-of-the-art agents trained for thousands of episodes in most scenarios and configurations. In addition, the best LLM agents perform similarly to human testers of the environment without any additional training process. This design highlights the potential of LLMs to efficiently address complex decision-making tasks within cybersecurity. Furthermore, we introduce a new network security environment named NetSecGame. The environment is designed to eventually support complex multi-agent scenarios within the network security domain. The proposed environment mimics real network attacks and is designed to be highly modular and adaptable for various scenarios., Comment: Under review. 10 pages plus appendices, 7 figures, 4 tables. Edit: fix e-mails and code repository

Honeypots are a well-known and widely used technology in the cybersecurity community, where it is assumed that placing honeypots in different geographical locations provides better visibility and increases effectiveness. However, how geolocation affects the usefulness of honeypots is not well-studied, especially for threat intelligence as early warning systems. This paper examines attack patterns in a large public dataset of geographically distributed honeypots by answering methodological questions and creating behavioural profiles of attackers. Results show that the location of honeypots helps identify attack patterns and build profiles for the attackers. We conclude that not all the intelligence collected from geographically distributed honeypots is equally valuable and that a good early warning system against resourceful attackers may be built with only two distributed honeypots and a production server.

Most network security datasets do not have comprehensive label assignment criteria, hindering the evaluation of the datasets, the training of models, the results obtained, the comparison with other methods, and the evaluation in real-life scenarios. There is no labeling ontology nor tools to help assign the labels, resulting in most analyzed datasets assigning labels in files or directory names. This paper addresses the problem of having a better labeling process by (i) reviewing the needs of stakeholders of the datasets, from creators to model users, (ii) presenting a new ontology of label assignment, (iii) presenting a new tool for assigning structured labels for Zeek network flows based on the ontology, and (iv) studying the differences between generating labels and consuming labels in real-life scenarios. We conclude that a process for structured label assignment is paramount for advancing research in network security and that the new ontology-based label assignation rules should be published as an artifact of every dataset.

Model stealing attacks have been successfully used in many machine learning domains, but there is little understanding of how these attacks work against models that perform malware detection. Malware detection and, in general, security domains have unique conditions. In particular, there are very strong requirements for low false positive rates (FPR). Antivirus products (AVs) that use machine learning are very complex systems to steal, malware binaries continually change, and the whole environment is adversarial by nature. This study evaluates active learning model stealing attacks against publicly available stand-alone machine learning malware classifiers and also against antivirus products. The study proposes a new neural network architecture for surrogate models (dualFFNN) and a new model stealing attack that combines transfer and active learning for surrogate creation (FFNN-TL). We achieved good surrogates of the stand-alone classifiers with up to 99\% agreement with the target models, using less than 4% of the original training dataset. Good surrogates of AV systems were also trained with up to 99% agreement and less than 4,000 queries. The study uses the best surrogates to generate adversarial malware to evade the target models, both stand-alone and AVs (with and without an internet connection). Results show that surrogate models can generate adversarial malware that evades the targets but with a lower success rate than directly using the target models to generate adversarial malware. Using surrogates, however, is still a good option since using the AVs for malware generation is highly time-consuming and easily detected when the AVs are connected to the internet., Comment: 20 pages, 10 figures, 8 tables. Accepted, please use the DOI/ journal for citations

Many activities related to cybercrime operations do not require much secrecy, such as developing websites or translating texts. This research provides indications that many users of a popular public internet marketing forum have connections to cybercrime. It does so by investigating the involvement in cybercrime of a population of users interested in internet marketing, both at a micro and macro scale. The research starts with a case study of three users confirmed to be involved in cybercrime and their use of the public forum where users share information about online advertising. It provides a first glimpse that some business with cybercrime connection is being conducted in the clear. The study then pans out to investigate the forum population's ties with cybercrime by finding crossover users, who are users from the public forum who also comment on cybercrime forums. The cybercrime forums on which they discuss are analyzed and crossover users' strength of participation is reported. Also, to assess if they represent a sub-group of the forum population, their posting behavior on the public forum is compared with that of non-crossover users. This blend of analyses shows that (i) a minimum of 7.2% of the public forum population are crossover users that have ties with cybercrime forums; (ii) their participation in cybercrime forums is limited; and (iii) their posting behavior is relatively indistinguishable from that of non-crossover users. This is the first study to formally quantify how users of an internet marketing public forum, a space for informal exchanges, have ties to cybercrime activities. We conclude that crossover users are a substantial part of the population in the public forum, and, even though they have thus far been overlooked, their aggregated effect in the ecosystem must be considered., Comment: 18 pages, 4 figures

Active Directory (AD) is a crucial element of large organizations, given its central role in managing access to resources. Since AD is used by all users in the organization, it is hard to detect attackers. We propose to generate and place fake users (honeyusers) in AD structures to help detect attacks. However, not any honeyuser will attract attackers. Our method generates honeyusers with a Variational Autoencoder that enriches the AD structure with well-positioned honeyusers. It first learns the embeddings of the original nodes and edges in the AD, then it uses a modified Bidirectional DAG-RNN to encode the parameters of the probability distribution of the latent space of node representations. Finally, it samples nodes from this distribution and uses an MLP to decide where the nodes are connected. The model was evaluated by the similarity of the generated AD with the original, by the positions of the new nodes, by the similarity with GraphRNN and finally by making real intruders attack the generated AD structure to see if they select the honeyusers. Results show that our machine learning model is good enough to generate well-placed honeyusers for existing AD structures so that intruders are lured into them., Comment: 2nd International Conference on Deep Learning Theory and Applications - DeLTA2021

Several encryption proposals for DNS have been presented since 2016, but their adoption was not comprehensively studied yet. This research measured the current adoption of DoH (DNS over HTTPS), DoT (DNS over TLS), and DoQ (DNS over QUIC) for five months at the beginning of 2021 by three different organizations with global coverage. By comparing the total values, amount of requests per user, and the seasonality of the traffic, it was possible to obtain the current adoption trends. Moreover, we actively scanned the Internet for still-unknown working DoH servers and we compared them with a novel curated list of well-known DoH servers. We conclude that despite growing in 2020, during the first five months of 2021 there was statistically significant evidence that the average amount of Internet traffic for DoH, DoT and DoQ remained stationary. However, we found that the amount of, still unknown and ready to use, DoH servers grew 4 times. These measurements suggest that even though the amount of encrypted DNS is currently not growing, there may probably be more connections soon to those unknown DoH servers for benign and malicious purposes., Comment: 16 pages, 10 figures

As machine learning becomes more widely used, the need to study its implications in security and privacy becomes more urgent. Although the body of work in privacy has been steadily growing over the past few years, research on the privacy aspects of machine learning has received less focus than the security aspects. Our contribution in this research is an analysis of more than 40 papers related to privacy attacks against machine learning that have been published during the past seven years. We propose an attack taxonomy, together with a threat model that allows the categorization of different attacks based on the adversarial knowledge, and the assets under attack. An initial exploration of the causes of privacy leaks is presented, as well as a detailed analysis of the different attacks. Finally, we present an overview of the most commonly proposed defenses and a discussion of the open problems and future directions identified during our analysis., Comment: Edit to add DOI. Accepted in ACM Computing Surveys, please cite the journal version

Domain Name Service is a trusted protocol made for name resolution, but during past years some approaches have been developed to use it for data transfer. DNS Tunneling is a method where data is encoded inside DNS queries, allowing information exchange through the DNS. This characteristic is attractive to hackers who exploit DNS Tunneling method to establish bidirectional communication with machines infected with malware with the objective of exfiltrating data or sending instructions in an obfuscated way. To detect these threats fast and accurately, the present work proposes a detection approach based on a Convolutional Neural Network (CNN) with a minimal architecture complexity. Due to the lack of quality datasets for evaluating DNS Tunneling connections, we also present a detailed construction and description of a novel dataset that contains DNS Tunneling domains generated with five well-known DNS tools. Despite its simple architecture, the resulting CNN model correctly detected more than 92% of total Tunneling domains with a false positive rate close to 0.8%.