Search

Our society increasingly benefits from Artificial Intelligence (AI). Unfortunately, more and more evidence shows that AI is also used for offensive purposes. Prior works have revealed various examples of use cases in which the deployment of AI can lead to violation of security and privacy objectives. No extant work, however, has been able to draw a holistic picture of the offensive potential of AI. In this SoK paper we seek to lay the ground for a systematic analysis of the heterogeneous capabilities of offensive AI. In particular we (i) account for AI risks to both humans and systems while (ii) consolidating and distilling knowledge from academic literature, expert opinions, industrial venues, as well as laypeople -- all of which being valuable sources of information on offensive AI. To enable alignment of such diverse sources of knowledge, we devise a common set of criteria reflecting essential technological factors related to offensive AI. With the help of such criteria, we systematically analyze: 95 research papers; 38 InfoSec briefings (from, e.g., BlackHat); the responses of a user study (N=549) entailing individuals with diverse backgrounds and expertise; and the opinion of 12 experts. Our contributions not only reveal concerning ways (some of which overlooked by prior work) in which AI can be offensively used today, but also represent a foothold to address this threat in the years to come., Comment: Systematization of Knowledge (SoK) paper. Accepted to the 3rd IEEE Conference on Secure and Trustworthy Machine Learning (SaTML'25)

Billions of individuals engage with Online Social Networks (OSN) daily. The owners of OSN try to meet the demands of their end-users while complying with business necessities. Such necessities may, however, lead to the adoption of restrictive data access policies that hinder research activities from "external" scientists -- who may, in turn, resort to other means (e.g., rely on static datasets) for their studies. Given the abundance of literature on OSN, we -- as academics -- should take a step back and reflect on what we have done so far, after having written thousands of papers on OSN. This is the first paper that provides a holistic outlook to the entire body of research that focused on OSN -- since the seminal work by Acquisti and Gross (2006). First, we search through over 1 million peer-reviewed publications, and derive 13,842 papers that focus on OSN: we organize the metadata of these works in the Minerva-OSN dataset, the first of its kind -- which we publicly release. Next, by analyzing Minerva-OSN, we provide factual evidence elucidating trends and aspects that deserve to be brought to light, such as the predominant focus on Twitter or the difficulty in obtaining OSN data. Finally, as a constructive step to guide future research, we carry out an expert survey (n=50) with established scientists in this field, and coalesce suggestions to improve the status quo such as an increased involvement of OSN owners. Our findings should inspire a reflection to "rescue" research on OSN. Doing so would improve the overall OSN ecosystem, benefiting both their owners and end-users and, hence, our society., Comment: Accepted at ICWSM 2025

We investigate the potential of using Large Language Models (LLM) to support process model creation in organizational contexts. Specifically, we carry out a case study wherein we develop and test an LLM-based chatbot, PRODIGY (PROcess moDellIng Guidance for You), in a multinational company, the Hilti Group. We are particularly interested in understanding how LLM can aid (human) modellers in creating process flow diagrams. To this purpose, we first conduct a preliminary user study (n=10) with professional process modellers from Hilti, inquiring for various pain-points they encounter in their daily routines. Then, we use their responses to design and implement PRODIGY. Finally, we evaluate PRODIGY by letting our user study's participants use PRODIGY, and then ask for their opinion on the pros and cons of PRODIGY. We coalesce our results in actionable takeaways. Through our research, we showcase the first practical application of LLM for process modelling in the real world, shedding light on how industries can leverage LLM to enhance their Business Process Management activities., Comment: 10 pages, 6 figures

Modern spacecraft are increasingly relying on machine learning (ML). However, physical equipment in space is subject to various natural hazards, such as radiation, which may inhibit the correct operation of computing devices. Despite plenty of evidence showing the damage that naturally-induced faults can cause to ML-related hardware, we observe that the effects of radiation on ML models for space applications are not well-studied. This is a problem: without understanding how ML models are affected by these natural phenomena, it is uncertain "where to start from" to develop radiation-tolerant ML software. As ML researchers, we attempt to tackle this dilemma. By partnering up with space-industry practitioners specialized in ML, we perform a reflective analysis of the state of the art. We provide factual evidence that prior work did not thoroughly examine the impact of natural hazards on ML models meant for spacecraft. Then, through a "negative result", we show that some existing open-source technologies can hardly be used by researchers to study the effects of radiation for some applications of ML in satellites. As a constructive step forward, we perform simple experiments showcasing how to leverage current frameworks to assess the robustness of practical ML models for cloud detection against radiation-induced faults. Our evaluation reveals that not all faults are as devastating as claimed by some prior work. By publicly releasing our resources, we provide a foothold -- usable by researchers without access to spacecraft -- for spearheading development of space-tolerant ML models.

Machine learning based phishing website detectors (ML-PWD) are a critical part of today's anti-phishing solutions in operation. Unfortunately, ML-PWD are prone to adversarial evasions, evidenced by both academic studies and analyses of real-world adversarial phishing webpages. However, existing works mostly focused on assessing adversarial phishing webpages against ML-PWD, while neglecting a crucial aspect: investigating whether they can deceive the actual target of phishing -- the end users. In this paper, we fill this gap by conducting two user studies (n=470) to examine how human users perceive adversarial phishing webpages, spanning both synthetically crafted ones (which we create by evading a state-of-the-art ML-PWD) as well as real adversarial webpages (taken from the wild Web) that bypassed a production-grade ML-PWD. Our findings confirm that adversarial phishing is a threat to both users and ML-PWD, since most adversarial phishing webpages have comparable effectiveness on users w.r.t. unperturbed ones. However, not all adversarial perturbations are equally effective. For example, those with added typos are significantly more noticeable to users, who tend to overlook perturbations of higher visual magnitude (such as replacing the background). We also show that users' self-reported frequency of visiting a brand's website has a statistically negative correlation with their phishing detection accuracy, which is likely caused by overconfidence. We release our resources.

Many domains now leverage the benefits of Machine Learning (ML), which promises solutions that can autonomously learn to solve complex tasks by training over some data. Unfortunately, in cyberthreat detection, high-quality data is hard to come by. Moreover, for some specific applications of ML, such data must be labeled by human operators. Many works "assume" that labeling is tough/challenging/costly in cyberthreat detection, thereby proposing solutions to address such a hurdle. Yet, we found no work that specifically addresses the process of labeling 'from the viewpoint of ML security practitioners'. This is a problem: to this date, it is still mostly unknown how labeling is done in practice -- thereby preventing one from pinpointing "what is needed" in the real world. In this paper, we take the first step to build a bridge between academic research and security practice in the context of data labeling. First, we reach out to five subject matter experts and carry out open interviews to identify pain points in their labeling routines. Then, by using our findings as a scaffold, we conduct a user study with 13 practitioners from large security companies, and ask detailed questions on subjects such as active learning, costs of labeling, and revision of labels. Finally, we perform proof-of-concept experiments addressing labeling-related aspects in cyberthreat detection that are sometimes overlooked in research. Altogether, our contributions and recommendations serve as a stepping stone to future endeavors aimed at improving the quality and robustness of ML-driven security systems. We release our resources.

The \emph{Markov numbers} are positive integers appearing as solutions to the Diophantine equation $x^2 + y^2 + z^2 = 3xyz$. These numbers are very well-studied and have many combinatorial properties, as well as being the source of the long-standing unicity conjecture. In 2018, \c{C}anak\c{c}{\i} and Schiffler showed that the Markov number $m_{\frac{a}{b}}$ is the number of perfect matchings of a certain snake graph corresponding to the Christoffel path from $(0,0)$ to $(a,b)$. Based on this correspondence, Schiffler in 2023 introduced two orderings on lattice paths. For any path $\omega$, associate a snake graph $\mathcal{G}(\omega)$ and a continued fraction $g(\omega)$. The ordering $<_M$ is given by the number of perfect matchings on $\mathcal{G}(\omega)$, and the ordering $<_L$ is given by the Lagrange number of $g(\omega)$. In this work, we settle two conjectures of Schiffler. First, we show that the path $\omega(a,b) = RR\cdots R UU \cdots U$ is the unique maximum over all lattice paths from $(0,0)$ to $(a,b)$ with respect to both orderings $<_M$ and $<_L$. We then use this result to prove that $\sup L(\omega)$ over all lattice paths is exactly $1+\sqrt5$., Comment: 11 pages, 2 figures

Recent times have witnessed the rise of anti-phishing schemes powered by deep learning (DL). In particular, logo-based phishing detectors rely on DL models from Computer Vision to identify logos of well-known brands on webpages, to detect malicious webpages that imitate a given brand. For instance, Siamese networks have demonstrated notable performance for these tasks, enabling the corresponding anti-phishing solutions to detect even "zero-day" phishing webpages. In this work, we take the next step of studying the robustness of logo-based phishing detectors against adversarial ML attacks. We propose a novel attack exploiting generative adversarial perturbations to craft "adversarial logos" that evade phishing detectors. We evaluate our attacks through: (i) experiments on datasets containing real logos, to evaluate the robustness of state-of-the-art phishing detectors; and (ii) user studies to gauge whether our adversarial logos can deceive human eyes. The results show that our proposed attack is capable of crafting perturbed logos subtle enough to evade various DL models-achieving an evasion rate of up to 95%. Moreover, users are not able to spot significant differences between generated adversarial logos and original ones., Comment: To appear in ESORICS 2023

Machine Learning (ML) has become a valuable asset to solve many real-world tasks. For Network Intrusion Detection (NID), however, scientific advances in ML are still seen with skepticism by practitioners. This disconnection is due to the intrinsically limited scope of research papers, many of which primarily aim to demonstrate new methods ``outperforming'' prior work -- oftentimes overlooking the practical implications for deploying the proposed solutions in real systems. Unfortunately, the value of ML for NID depends on a plethora of factors, such as hardware, that are often neglected in scientific literature. This paper aims to reduce the practitioners' skepticism towards ML for NID by "changing" the evaluation methodology adopted in research. After elucidating which "factors" influence the operational deployment of ML in NID, we propose the notion of "pragmatic assessment", which enable practitioners to gauge the real value of ML methods for NID. Then, we show that the state-of-research hardly allows one to estimate the value of ML for NID. As a constructive step forward, we carry out a pragmatic assessment. We re-assess existing ML methods for NID, focusing on the classification of malicious network traffic, and consider: hundreds of configuration settings; diverse adversarial scenarios; and four hardware platforms. Our large and reproducible evaluations enable estimating the quality of ML for NID. We also validate our claims through a user-study with security practitioners.

This article presents a new method for estimating the electron temperature of the Protosphera's screw pinch. The temperature radial profile is obtained by a self-consistent modeling of a 1D MHD equilibrium along with a 0D power balance of the plasma column, given measurements and estimates of the axial pinch plasma current, of the plasma rotational frequency and, at the equatorial plane, of the electron density radial profile, of the edge poloidal magnetic field, of the edge electron temperature and of the neutrals pressure in the vacuum vessel. The plasma is considered in equilibrium with its neutral phase and in constant rotation. A MATLAB code has been developed with the aim of estimating the MHD radial equilibrium profiles, the thermodynamic plasma state and the neutrals profile. The numerical estimates are compared with available experimental data showing a good agreement., Comment: 4 pages, 6 figures, 1 table, research presented to the "6th ICFDT"

Recent years have seen a proliferation of research on adversarial machine learning. Numerous papers demonstrate powerful algorithmic attacks against a wide variety of machine learning (ML) models, and numerous other papers propose defenses that can withstand most attacks. However, abundant real-world evidence suggests that actual attackers use simple tactics to subvert ML-driven systems, and as a result security practitioners have not prioritized adversarial ML defenses. Motivated by the apparent gap between researchers and practitioners, this position paper aims to bridge the two domains. We first present three real-world case studies from which we can glean practical insights unknown or neglected in research. Next we analyze all adversarial ML papers recently published in top security conferences, highlighting positive trends and blind spots. Finally, we state positions on precise and cost-driven threat modeling, collaboration between industry and academia, and reproducible research. We believe that our positions, if adopted, will increase the real-world impact of future endeavours in adversarial ML, bringing both researchers and practitioners closer to their shared goal of improving the security of ML systems.

Although machine learning based algorithms have been extensively used for detecting phishing websites, there has been relatively little work on how adversaries may attack such "phishing detectors" (PDs for short). In this paper, we propose a set of Gray-Box attacks on PDs that an adversary may use which vary depending on the knowledge that he has about the PD. We show that these attacks severely degrade the effectiveness of several existing PDs. We then propose the concept of operation chains that iteratively map an original set of features to a new set of features and develop the "Protective Operation Chain" (POC for short) algorithm. POC leverages the combination of random feature selection and feature mappings in order to increase the attacker's uncertainty about the target PD. Using 3 existing publicly available datasets plus a fourth that we have created and will release upon the publication of this paper, we show that POC is more robust to these attacks than past competing work, while preserving predictive performance when no adversarial attacks are present. Moreover, POC is robust to attacks on 13 different classifiers, not just one. These results are shown to be statistically significant at the p < 0.001 level.

The Smart Grid (SG) is a cornerstone of modern society, providing the energy required to sustain billions of lives and thousands of industries. Unfortunately, as one of the most critical infrastructures of our World, the SG is an attractive target for attackers. The problem is aggravated by the increasing adoption of digitalisation, which further increases the SG's exposure to cyberthreats. Successful exploitation of such exposure leads to entire countries being paralysed, which is an unacceptable -- but ultimately inescapable -- risk. This paper aims to mitigate this risk by elucidating the perspective of real practitioners on the cybersecurity of the SG. We interviewed 18 entities, operating in diverse countries in Europe and covering all domains of the SG -- from energy generation, to its delivery. Our analysis highlights a stark contrast between (a)research and practice, but also between (b) public and private entities. For instance: some threats appear to be much less dangerous than what is claimed in related papers; some technological paradigms have dubious utility for practitioners, but are actively promoted by literature; finally, practitioners may either under- or over-estimate their own cybersecurity capabilities. We derive four takeaways that enable future endeavours to improve the overall cybersecurity in the SG. We conjecture that most of the problems are due to an improper communication between researchers, practitioners and regulatory bodies -- which, despite sharing a common goal, tend to neglect the viewpoint of the other `spheres'.

Existing literature on adversarial Machine Learning (ML) focuses either on showing attacks that break every ML model, or defenses that withstand most attacks. Unfortunately, little consideration is given to the actual feasibility of the attack or the defense. Moreover, adversarial samples are often crafted in the "feature-space", making the corresponding evaluations of questionable value. Simply put, the current situation does not allow to estimate the actual threat posed by adversarial attacks, leading to a lack of secure ML systems. We aim to clarify such confusion in this paper. By considering the application of ML for Phishing Website Detection (PWD), we formalize the "evasion-space" in which an adversarial perturbation can be introduced to fool a ML-PWD -- demonstrating that even perturbations in the "feature-space" are useful. Then, we propose a realistic threat model describing evasion attacks against ML-PWD that are cheap to stage, and hence intrinsically more attractive for real phishers. After that, we perform the first statistically validated assessment of state-of-the-art ML-PWD against 12 evasion attacks. Our evaluation shows (i) the true efficacy of evasion attempts that are more likely to occur; and (ii) the impact of perturbations crafted in different evasion-spaces. Our realistic evasion attempts induce a statistically significant degradation (3-10% at p<0.05), and their cheap cost makes them a subtle threat. Notably, however, some ML-PWD are immune to our most realistic attacks (p=0.22). Finally, as an additional contribution of this journal publication, we are the first to consider the intriguing case wherein an attacker introduces perturbations in multiple evasion-spaces at the same time. These new results show that simultaneously applying perturbations in the problem- and feature-space can cause a drop in the detection rate from 0.95 to 0.

Did you know that over 70 million of Dota2 players have their in-game data freely accessible? What if such data is used in malicious ways? This paper is the first to investigate such a problem. Motivated by the widespread popularity of video games, we propose the first threat model for Attribute Inference Attacks (AIA) in the Dota2 context. We explain how (and why) attackers can exploit the abundant public data in the Dota2 ecosystem to infer private information about its players. Due to lack of concrete evidence on the efficacy of our AIA, we empirically prove and assess their impact in reality. By conducting an extensive survey on $\sim$500 Dota2 players spanning over 26k matches, we verify whether a correlation exists between a player's Dota2 activity and their real-life. Then, after finding such a link ($p$ < 0.01 and $\rho$ > 0.3), we ethically perform diverse AIA. We leverage the capabilities of machine learning to infer real-life attributes of the respondents of our survey by using their publicly available in-game data. Our results show that, by applyingdomain expertise, some AIA can reach up to 98% precision and over 90% accuracy. This paper hence raises the alarm on a subtle, but concrete threat that can potentially affect the entire competitive gaming landscape. We alerted the developers of Dota2.

Fifth Generation (5G) networks must support billions of heterogeneous devices while guaranteeing optimal Quality of Service (QoS). Such requirements are impossible to meet with human effort alone, and Machine Learning (ML) represents a core asset in 5G. ML, however, is known to be vulnerable to adversarial examples; moreover, as our paper will show, the 5G context is exposed to a yet another type of adversarial ML attacks that cannot be formalized with existing threat models. Proactive assessment of such risks is also challenging due to the lack of ML-powered 5G equipment available for adversarial ML research. To tackle these problems, we propose a novel adversarial ML threat model that is particularly suited to 5G scenarios, and is agnostic to the precise function solved by ML. In contrast to existing ML threat models, our attacks do not require any compromise of the target 5G system while still being viable due to the QoS guarantees and the open nature of 5G networks. Furthermore, we propose an original framework for realistic ML security assessments based on public data. We proactively evaluate our threat model on 6 applications of ML envisioned in 5G. Our attacks affect both the training and the inference stages, can degrade the performance of state-of-the-art ML systems, and have a lower entry barrier than previous attacks.

Machine Learning (ML) represents a pivotal technology for current and future information systems, and many domains already leverage the capabilities of ML. However, deployment of ML in cybersecurity is still at an early stage, revealing a significant discrepancy between research and practice. Such discrepancy has its root cause in the current state-of-the-art, which does not allow to identify the role of ML in cybersecurity. The full potential of ML will never be unleashed unless its pros and cons are understood by a broad audience. This paper is the first attempt to provide a holistic understanding of the role of ML in the entire cybersecurity domain -- to any potential reader with an interest in this topic. We highlight the advantages of ML with respect to human-driven detection methods, as well as the additional tasks that can be addressed by ML in cybersecurity. Moreover, we elucidate various intrinsic problems affecting real ML deployments in cybersecurity. Finally, we present how various stakeholders can contribute to future developments of ML in cybersecurity, which is essential for further progress in this field. Our contributions are complemented with two real case studies describing industrial applications of ML as defense against cyber-threats.

Machine learning (ML) has become an important paradigm for cyberthreat detection (CTD) in the recent years. A substantial research effort has been invested in the development of specialized algorithms for CTD tasks. From the operational perspective, however, the progress of ML-based CTD is hindered by the difficulty in obtaining the large sets of labelled data to train ML detectors. A potential solution to this problem are semisupervised learning (SsL) methods, which combine small labelled datasets with large amounts of unlabelled data. This paper is aimed at systematization of existing work on SsL for CTD and, in particular, on understanding the utility of unlabelled data in such systems. To this end, we analyze the cost of labelling in various CTD tasks and develop a formal cost model for SsL in this context. Building on this foundation, we formalize a set of requirements for evaluation of SsL methods, which elucidates the contribution of unlabelled data. We review the state-of-the-art and observe that no previous work meets such requirements. To address this problem, we propose a framework for assessing the benefits of unlabelled data in SsL. We showcase an application of this framework by performing the first benchmark evaluation that highlights the tradeoffs of 9 existing SsL methods on 9 public datasets. Our findings verify that, in some cases, unlabelled data provides a small, but statistically significant, performance gain. This paper highlights that SsL in CTD has a lot of room for improvement, which should stimulate future research in this field.

We propose to generate adversarial samples by modifying activations of upper layers encoding semantically meaningful concepts. The original sample is shifted towards a target sample, yielding an adversarial sample, by using the modified activations to reconstruct the original sample. A human might (and possibly should) notice differences between the original and the adversarial sample. Depending on the attacker-provided constraints, an adversarial sample can exhibit subtle differences or appear like a "forged" sample from another class. Our approach and goal are in stark contrast to common attacks involving perturbations of single pixels that are not recognizable by humans. Our approach is relevant in, e.g., multi-stage processing of inputs, where both humans and machines are involved in decision-making because invisible perturbations will not fool a human. Our evaluation focuses on deep neural networks. We also show the transferability of our adversarial examples among networks., Comment: Accepted at IEEE Symposium on Security and Privacy (S&P) Workshop on Deep Learning and Security, 2022

Enhancing Network Intrusion Detection Systems (NIDS) with supervised Machine Learning (ML) is tough. ML-NIDS must be trained and evaluated, operations requiring data where benign and malicious samples are clearly labelled. Such labels demand costly expert knowledge, resulting in a lack of real deployments, as well as on papers always relying on the same outdated data. The situation improved recently, as some efforts disclosed their labelled datasets. However, most past works used such datasets just as a 'yet another' testbed, overlooking the added potential provided by such availability. In contrast, we promote using such existing labelled data to cross-evaluate ML-NIDS. Such approach received only limited attention and, due to its complexity, requires a dedicated treatment. We hence propose the first cross-evaluation model. Our model highlights the broader range of realistic use-cases that can be assessed via cross-evaluations, allowing the discovery of still unknown qualities of state-of-the-art ML-NIDS. For instance, their detection surface can be extended--at no additional labelling cost. However, conducting such cross-evaluations is challenging. Hence, we propose the first framework, XeNIDS, for reliable cross-evaluations based on Network Flows. By using XeNIDS on six well-known datasets, we demonstrate the concealed potential, but also the risks, of cross-evaluations of ML-NIDS.

The incremental diffusion of machine learning algorithms in supporting cybersecurity is creating novel defensive opportunities but also new types of risks. Multiple researches have shown that machine learning methods are vulnerable to adversarial attacks that create tiny perturbations aimed at decreasing the effectiveness of detecting threats. We observe that existing literature assumes threat models that are inappropriate for realistic cybersecurity scenarios because they consider opponents with complete knowledge about the cyber detector or that can freely interact with the target systems. By focusing on Network Intrusion Detection Systems based on machine learning, we identify and model the real capabilities and circumstances required by attackers to carry out feasible and successful adversarial attacks. We then apply our model to several adversarial attacks proposed in literature and highlight the limits and merits that can result in actual adversarial attacks. The contributions of this paper can help hardening defensive systems by letting cyber defenders address the most critical and real issues, and can benefit researchers by allowing them to devise novel forms of adversarial attacks based on realistic threat models.

Recent advances in deep learning renewed the research interests in machine learning for Network Intrusion Detection Systems (NIDS). Specifically, attention has been given to sequential learning models, due to their ability to extract the temporal characteristics of Network traffic Flows (NetFlows), and use them for NIDS tasks. However, the applications of these sequential models often consist of transferring and adapting methodologies directly from other fields, without an in-depth investigation on how to leverage the specific circumstances of cybersecurity scenarios; moreover, there is a lack of comprehensive studies on sequential models that rely on NetFlow data, which presents significant advantages over traditional full packet captures. We tackle this problem in this paper. We propose a detailed methodology to extract temporal sequences of NetFlows that denote patterns of malicious activities. Then, we apply this methodology to compare the efficacy of sequential learning models against traditional static learning models. In particular, we perform a fair comparison of a `sequential' Long Short-Term Memory (LSTM) against a `static' Feedforward Neural Networks (FNN) in distinct environments represented by two well-known datasets for NIDS: the CICIDS2017 and the CTU13. Our results highlight that LSTM achieves comparable performance to FNN in the CICIDS2017 with over 99.5\% F1-score; while obtaining superior performance in the CTU13, with 95.7\% F1-score against 91.5\%. This paper thus paves the way to future applications of sequential learning models for NIDS.