The development of generative algorithms in the last few years, including large language models, imposes increasingly high requirements in the field of data security and protection. Vulnerabilities of information systems related to the generation of inappropriate information are becoming an increasingly serious challenge. They can lead to negative consequences, such as misinformation, creation of fake news or disclosure of sensitive data. This paper presents the architecture of a system designed to identify vulnerabilities in large language models and proposes an approach for vulnerability detection based on prompt engineering. The idea of the approach lies in making certain requests to large language models, the execution of which can help to use the algorithm for illegal purposes. The authors provide a detailed description of the system prototype for automating vulnerability detection, named LLMSecurityTester. Additionally, preliminary results of experiments on various models and datasets are presented, demonstrating the applicability of the solution.

The objective of this paper is to explore an effective approach for detecting anomalies in cryptocurrency transactions using artificial intelligence techniques. The proposed method is compared with existing approaches for identifying illegal transactions within cryptocurrency networks. This research systematically analyzes various data processing methods applicable to the problem of detecting illicit transactions, as well as reviewing relevant studies in cryptocurrency transaction analysis, data visualization, and computer vision-based anomaly detection. Experimental evaluations on a cryptocurrency transaction dataset show significant insights, with results benchmarked against those obtained in prior studies on the same dataset (BABD). The findings suggest that the proposed method, utilizing statistical graph characteristics, deep and ensemble learning technologies, enhances the accuracy of identifying fraudulent transactions. This approach could be instrumental in developing software solutions for monitoring cryptocurrency transactions, potentially aiding in digital forensics and cybersecurity applications. The findings may benefit researchers in computer security and developers focused on information security systems.

This paper describes an original approach of classification with active learning for inappropriate information detection and its application for the text posts from the VKontakte social network. The novelty of the approach lies in the constantly growing dataset, while the classifiers training process takes place during the operator's work. The approach works with texts of any size and content and applicable for Russian social networks. The research contribution lies in the original approach for inappropriate information detection, while practical significance lies in the automation of routine tasks to reduce the burden on specialists in the area of protection from information. Experimental evaluation of the approach is focused on its iterative retraining part. For the experiment, text posts of different topics from the VKontakte social network were collected and labeled. After that, we have evaluated F-measure and ROC-AUC metrics for classifiers trained on random subsamples of different sizes and different topics. Moreover, the advantages and disadvantages of the approach, as well as future work directions, were indicated.

The paper considers the task of bot detection in social networks. Study is focused on the case when the account is closed by the privacy settings, and the bot needs to be identified by the friend list. The paper proposes a solution that is based on machine learning and statistical methods. Social network VKontakte is used as a data source. The paper provides the review of data, that one needs to get from the social network for bot detection in the case when the profile is closed by privacy settings. The paper includes a description of features extraction from VKontakte social network and extracting complexity evaluation; description of features construction using statistics, Benford's law and Gini index. The paper describes the experiment. To collect data for training, we collect bots and real users datasets. To collect bots we made fake groups and bought bots of different quality for them from 3 different companies. We performed two series of experiments. In the first series, all the features were used to train and evaluate the classifiers. In the second series of experiments, the features were preliminary examined for the presence of strong correlation between them. The results demonstrated the feasibility of rather highaccuracy private account bot detection by means of unsophisticated off-the-shelf algorithms in combination with data scaling. The Random Forest Classifier yields the best results, with an ROC AUC more than 0.9 and FPR less than 0.3. In paper we also discuss the limitations of the experimental part, and plans for future research.

The development and evaluation of security solutions for autonomous vehicles is a challenging task. Many researchers have no access to real vehicles to implement and test their solutions. In addition, vehicle E/E architectures of different brands or even model series of one car manufacturer differ significantly. Also, vehicles may be the source of physical hazards, e.g., an exploding airbag. To enable researchers to develop, implement, and evaluate new security solutions for autonomous vehicles, we propose a new security evaluation platform called SEPAD and a dedicated development process for testing security mechanisms with it. SEPAD allows to model realistic E/E architectures where the developed security solutions can be integrated and evaluated without causing safety risks for the researcher or other road users.

The paper considers the task of bot detection in social networks. It checks the hypothesis that bots break Benford’s law much more often than users, so one can identify them. A bot detection approach is proposed based on experiments where the test results for bot datasets of different classes and real-user datasets of different communities are evaluated and compared. The experiments show that automatically controlled bots possibly can be identified by disagreement with Benford’s law, while human-orchestrated bots are not.

From information security point of view embedded devices are the elements of complex systems operating in a potentially hostile environment. Therefore development of embedded devices is a complex task that often requires expert solutions. The complexity of the task of developing secure embedded devices is caused by various types of threats and attacks that may affect the device, as well as that in practice security of embedded devices is usually considered at the final stage of the development process in the form of adding additional security features. The paper proposes a design technique and its application that will facilitate development of secure and energy-efficient embedded devices. The technique organizes the search for the best combinations of security components on the basis of solving an optimization problem. The efficiency of the proposed technique is demonstrated by development of a room perimeter protection system.