Abstract: A rootkit detection system and method analyzes memory dumps to determine connections between intercepted system driver operations requested by unknown files and changes in system memory before and after those operations. Memory dump differences and I/O buffers are analyzed with machine learning models to identify clustered features associated with rootkits.

Abstract: Systems and methods for trusted storage of encrypted data onto a cloud storage server are disclosed. The system includes a client device and the cloud storage server. The client device includes a processing unit, a client agent, and a trusted data loading device. The client agent has no access to the network and is implemented within the client device. The client agent receives data and requests to upload the data from the processing unit. The client agent encrypts the data and wraps the data into an encrypted data container.

Abstract: A system and method for malware classification using machine learning models trained using synthesized feature sets based on features extracted from samples of known malicious objects and known safe objects. The synthesized feature sets act as virtual samples for training a machine learning classifier to recognize new objects in the wild that are likely to be malicious.

Abstract: A system and method for malware detection uses static and dynamic analysis to train a machine learning model. At the training step, static and dynamic features are extracted from training datasets and used to train a malware classification model. The malware classification model is used to classify unknown files based on verdicts from both static and dynamic models.

Abstract: A system and method for malware detection uses static and dynamic analysis to augment a machine learning model. At the training step, static and dynamic features are extracted from training datasets and used to train a malware classification model. The malware classification model is used to classify unknown files based on verdicts from both static and dynamic models.

Abstract: A distributed agent for backup and restoration of virtual machines collects backup data and meta-data. The distributed agent includes an agent inside a virtual machine and an agent outside the virtual machine. The two kinds of agents communicate with each other to collect data of different types used to backup and restore virtual machines.

Abstract: Disclosed herein are systems and method for restoring a clean backup after a malware attack. In one aspect, a method forms a list of files that are of a plurality of designated file types that can be infected by malicious software. The method performs one or more snapshots of the files according to a predetermined schedule over a predetermined period of time and performs one or more backups. The method determines that a malware attack is being carried out on the computing device and generates a list of dangerous objects that spread the malware attack. The method compares the list of dangerous objects with the one or more snapshots to determine when the malware attack occurred. The method identifies a clean backup that was created most recently before the malware attack as compared to other backups and recovers data for the computing device from the clean backup.

Abstract: Systems and methods for file sharing over secure connections.

Abstract: Disclosed herein are systems and method for anti-malware scanning, including identifying a plurality of objects in a backup archive that is connected to a first network comprising a plurality of computing devices; scanning the plurality of objects in the backup archive to generate a whitelist indicating a subset of the plurality of objects that do not need to be scanned at a subsequent time; performing, using the whitelist, a first malware scan in a computing device of the plurality of computing devices; detecting that the computing device has left the first network to join a second network; and performing a second malware scan on the computing device, wherein the second malware scan uses a different whitelist of the second network, and wherein the second malware scan comprises scanning a first object that is not in the different whitelist and was not scanned in the first malware scan.

Abstract: Disclosed herein are systems and method for detecting passwords vulnerable to compromise. In one exemplary aspect, a method comprises identifying a plurality of files in at least one storage device of an organization. For each respective file in the plurality of files, in response to determining that the respective file type is in the database of vulnerable file types, the method comprises parsing text in the respective file and identifying, for the respective file, at least one demographic associated with the organization. The method further comprises retrieving dictionaries and expressions specific to the at least one demographic and determining the text in the respective file comprises a password using the retrieved dictionaries and expressions of the at least one demographic. In response to determining that the text comprises the password, the method comprises generating a security alert for an administrator of the storage device.

Abstract: Disclosed herein are systems and methods for verifying user activity based on behavioral models. In an exemplary aspect, a method may include receiving and parsing sensor data from at least one sensor in an environment to determine a first identifier of a person that is not authorized to access data via a computing device. The method may include intercepting, on the computing device, a data access request including a second identifier of a user that is authorized to access the data via the computing device. The method may include verifying whether the data access request is from the user authorized to access data by determining whether a chain of events involving the first person and the user corresponds to a behavioral model indicative of malicious activity. Based on the verification, the data access request is either granted or denied.

Abstract: Described herein are systems and methods for controlling access to a protected resource based on various criteria. In one exemplary aspect, a method comprises designating a plurality of program data installed on a computing system as protected program data; intercepting, by a kernel mode driver, a request from an untrusted application executing on the computing system to alter at least one of the protected program data; classifying, by a self-defense service, the untrusted application as a malicious application based on the intercepted request and information related to the untrusted application; and responsive to classifying the untrusted application as a malicious application, denying, by the kernel mode driver, access to the at least one of the protected program data.

Abstract: Disclosed herein are systems and methods for preventing anti-forensics actions. In one exemplary aspect, a method may identify a suspicious object from a plurality of objects on a computing device and monitor actions performed by the suspicious object. The method may intercept a first command by the suspicious object to create and/or modify a digital artifact on the computing device and subsequent to intercepting the first command, intercept a second command by the suspicious object to delete at least one of the suspicious object and the digital artifact. In response to intercepting both the first command to create and/or modify the digital artifact and the second command to delete at least one of the suspicious object and the digital artifact, the method may block the second command, and may store the suspicious object and the digital artifact in a digital repository.

Abstract: Disclosed herein are systems and methods for continuous user authentication during access of a digital service. In an exemplary aspect, a continuous authentication module may receive, at a computing device, initial authentication credentials of the user. The initial authentication credentials enable access to a service via the computing device. While the service is being accessed, the continuous authentication module may continuously monitor whether an unauthorized user has replaced the user in accessing the service by comparing usage attributes of the service with historic usage attributes associated with the user. In response to determining that the unauthorized user has replaced the user, the continuous authentication module may cease the access to the service via the computing device.

Abstract: Disclosed herein are systems and method for verifying user identity. In one exemplary aspect, a method comprises receiving sensor data from at least one sensor and parsing the sensor data to determine a plurality of identifiers of a user authorized to access data via a computing device. The method comprises generating, for each identifier, an event including the user. The method comprises intercepting, on the computing device, a data access request including an identifier of the user. The method comprises verifying whether the data access request is from the user authorized to access data by generating a chain of events including the user for a period of time preceding the data access request, determining a deviation of the chain of events from a target chain of events, and in response to determining the deviation is less than a threshold deviation value, granting the data access request.

Abstract: Disclosed herein are systems and method for backing up data in a load-balanced clustered environment. A clustered resource to be backed up is selected, wherein the clustered resource is stored on a common storage system and operated on by a cluster-aware application executing on at least a first node and a second node of a computing cluster. A load-balanced application may migrate the clustered resource from the first node with a high-load consumption to the second node with low-load consumption. A list of changes made by both nodes are received and merged. A backup agent then generates a consistent incremental backup using data retrieved from the common storage system according to the merged list of changes to the clustered resource.

Abstract: Disclosed herein are systems and methods for pattern-based QoS violation prediction. In one exemplary aspect, a method may comprise identifying a service on a computing device that is connected to a plurality of client devices and determining a plurality of micro-services comprised in the identified service. The method may comprise parsing access information to detect that a first client device is accessing a micro-service. The method may comprise determining, for a first period of time, QoS evaluation parameters for the access between the micro-service and the first client device. The method may comprise identifying changes in the QoS evaluation parameters within the first period of time, detecting a predetermined QoS violation pattern, and executing a QoS action based on the predetermined QoS violation pattern.

Abstract: Disclosed herein are systems and method for performing recovery using a backup image. In one exemplary aspect, a method comprises scanning a plurality of files on one or more storage devices of a computing device. The method may determine a first set of files from the plurality of files that will be used during recovery of the one or more storage devices, and tag a second set of files that will not be used during recovery. The method may copy the second set of files that have been tagged to an external storage device, and may store the first set of files in a backup image for the computing device (excluding the tagged second set of files from the backup image). The method may add, to the backup image, a respective link to each of the tagged second set of files in the external storage device.

Abstract: Disclosed herein are systems and method for subduing target individuals using unmanned aerial vehicles (UAVs) comprises deploying one or more UAVs to a location of an individual. The method obtains information about the individual from the one or more UAVs and external sources. The method assigns an aggression factor to the individual based on the obtained information. In response to determining that the aggression factor is greater than an aggression threshold, the method prepares a neutralization action designed to reduce the aggression factor of the individual by: identifying one or more exception conditions of the individual, and selecting, from a plurality of neutralization actions that the one or more UAVs are capable of performing, a neutralization action based on the one or more exception conditions. The method then instructs the one or more UAVs to perform the selected neutralization action on the hostile individuals.

Abstract: A system and method for detecting malware using hierarchical clustering analysis. Unknown files classified by clustering and in view of known malicious and known safe files. Machine learning models and detection rules are used to enhance classification accuracy.