Abstract: Methods, systems, and computer readable media for network security are described. In some implementations, security tasks and roles can be allocated between an endpoint device and a firewall device based on tag information sent from the endpoint, the tag information including one or more characteristics of a traffic flow, information of resource availability, and/or reputation of a process associated with a traffic flow.

Abstract: A collection of documents or other files and the like within an enterprise network are labelled according to an enterprise document classification scheme, and then a recognition model such as a neural network or other machine learning model can be used to automatically label other files throughout the enterprise network. In this manner, documents and the like throughout an enterprise can be automatically identified and managed according to features such as confidentiality, sensitivity, security risk, business value, and so forth.

Abstract: A computer-implemented method includes generating behavior patterns based on historical behavior of a plurality of emails. The method further includes receiving an email message from a sender, wherein the email message is withheld from delivery to a recipient. The method further includes extracting a plurality of features from the email message. The method further includes determining whether content of the email message matches at least one criterion for suspicious content. The method further includes determining a reputation score associated with the sender based on a comparison of the extracted features with the behavior patterns, wherein the extracted features include an identity of the sender. The method further includes responsive to the content of the email message not matching the at least one criterion for suspicious content and the reputation score meeting a reputation threshold, delivering the email message to the recipient.

Abstract: An example method can include, obtaining information about a machine learning (ML) model configured to detect malicious activity on a computer system, wherein the information includes one or more of a model type of the ML model, an output type of the ML model, or a type of malicious activity that the model is trained to detect, receiving a training dataset, wherein the training dataset includes a plurality of unlabeled examples, generating an additional dataset based on the training dataset using a generative model, wherein the additional dataset includes a plurality of additional unlabeled examples, and, training the machine learning model to generate labels for each example in the training dataset and the additional dataset, using a combination of the training dataset and the additional dataset, wherein the training includes adjusting one or more parameters of the machine learning model based on accuracy of the generated labels.

Abstract: A file integrity monitoring system supports monitoring of system-critical, enterprise-critical and user-critical data by reporting events to a threat management facility in response to changes in certain files, folders, registry keys and registry values of the computing environment in which the system is operating and/or monitoring. The file integrity monitoring system may dynamically create, adapt and apply context-based rules to improve the sensitivity and relevance of reported events to undesirable changes in the data footprint of a monitored device.

Abstract: In embodiments of the present invention improved capabilities are described for the steps of receiving an indication that a computer facility has access to a secure data store, causing a security parameter of a storage medium local to the computer facility to be assessed, determining if the security parameter is compliant with a security policy relating to computer access of the remote secure data store, and in response to an indication that the security parameter is non-compliant, cause the computer facility to implement an action to prevent further dissemination of information, to disable access to network communications, to implement an action to prevent further dissemination of information, and the like.

Abstract: Implementations generally relate methods, systems, and computer readable media for providing automatic access point registration. In some implementations, a method includes receiving an indication of automatic device onboarding activation. The method further includes receiving a selection of one or more reference devices. The method further includes determining one or more detectable devices of the one or more candidate devices to be onboarded that are detectable by at least one of the one or more reference devices. The method further includes obtaining one or more automatic configuration parameters from one or more of the reference devices. The method further includes configuring one or more of the detectable devices to be onboarded with the one or more automatic configuration parameters.

Abstract: Methods, systems, and computer readable media for rogue device detection are described. A method may include identifying a device type of a device transmitting data over a network and obtaining one or more uniform resource locators (URLs) from the data, where the one or more URLs form a portion of a request transmitted over the network by the device. The method can also include programmatically analyzing the data to determine a pattern of network data within a given time period. The method can further include determining that the device is a rogue device if the pattern of network data deviates from a baseline pattern of the device type, or at least one of the one or more URLs matches one or more rogue URL criteria. The method can also include taking an action in response to determining the device is a rogue device to improve security of the network.

Abstract: An automated system attempts to characterize code as safe or unsafe. For intermediate code samples not placed with sufficient confidence in either category, human-readable analysis is automatically generated to assist a human reviewer in reaching a final disposition. For example, a random forest over human-interpretable features may be created and used to identify suspicious features in a manner that is understandable to, and actionable by, a human reviewer. Similarly, a k-nearest neighbor algorithm may be used to identify similar samples of known safe and unsafe code based on a model for, e.g., a file path, a URL, an executable, and so forth. Similar code may then be displayed (with other information) to a user for evaluation in a user interface. This comparative information can improve the speed and accuracy of human interventions by providing richer context for human review of potential threats.

Abstract: Attachments or other documents can be transmitted to a sandbox environment where they can be concurrently opened for remote preview from an endpoint and scanned for possible malware. A gateway or other intermediate network element may enforce this process by replacing attachments, for example, in incoming electronic mail communications, with links to a document preview hosted in the sandbox environment.

Abstract: An endpoint in a network periodically generates a heartbeat encoding health state information and transmits this heartbeat to other network entities. Recipients of the heartbeat may use the health state information to independently make decisions about communications with the source endpoint, for example, by isolating the endpoint to prevent further communications with other devices sharing the network with the endpoint. Isolation may be coordinated by a firewall or gateway for the network, or independently by other endpoints that receive a notification of the compromised health state.

Abstract: A threat management system provides a collection of queries for investigating security issues within an enterprise. Useful inferences are drawn about the value of different queries, and about the security posture of the enterprise, by monitoring contextual activity such as the popularity and context of query usage, patterns of end user modification to queries, and post-query activity.

Abstract: An endpoint coupled in a communicating relationship with an enterprise network may include a data recorder configured to store an event stream of data indicating events on the endpoint including types of changes to computing objects, a filter configured to locally process the event stream into a filtered event stream including a subset of types of changes to the computing objects, and a local security agent. The local security agent may be configured to transmit the filtered event stream to a threat management facility, respond to a filter adjustment from the threat management facility by adjusting the filter to modify the subset of types of changes included in the filtered event stream, and respond to a query from the threat management facility by retrieving data stored in the data recorder over a time window before the query and excluded from the filtered event stream.

Abstract: A variety of techniques are disclosed for detection of advanced persistent threats and similar malware. In one aspect, the detection of certain network traffic at a gateway is used to trigger a query of an originating endpoint, which can use internal logs to identify a local process that is sourcing the network traffic. In another aspect, an endpoint is configured to periodically generate and transmit a secure heartbeat, so that an interruption of the heartbeat can be used to signal the possible presence of malware. In another aspect, other information such as local and global reputation information is used to provide context for more accurate malware detection.

Abstract: In a threat management platform, a number of endpoints log events in an event data recorder. A local agent filters this data and feeds a filtered data stream to a central threat management facility. The central threat management facility can locally or globally tune filtering by local agents based on the current data stream, and can query local event data recorders for additional information where necessary or helpful in threat detection or forensic analysis. The central threat management facility also stores and deploys a number of security tools such as a web-based user interface supported by machine learning models to identify potential threats requiring human intervention and other models to provide human-readable context for evaluating potential threats.

Abstract: A firewall uses information about an application that originates a network request to determine whether and how to forward the request over a network. The firewall may more generally rely on the identity of the originating application, the security state of the originating application, the security state of the endpoint, and any other information that might provide an indication of malicious activity, to make routing and forwarding decisions for endpoint-originated network traffic.

Abstract: In embodiments of the present invention, a framework for an extensible, file-based security system is described for determining an appropriate application, application environment, and/or access or security control measure based at least in part on a file's reputation. In response to the selection of a file, an application controller may be used to select a software application from two or more software applications to open the selected file, based at least in part on the selected file's reputation. If launched, a software application may be configured to open the file in an environment, such as a virtual machine, quarantined environment, and the like, that is appropriate for the file based at least in part on the reputation information. A software application may be a secure software application configured to manage secure files, or an insecure software application configured to manage insecure files.

Abstract: A honeypot file is cryptographically secured with a cryptographic key. The key, or related key material, is then placed on a central keystore and the file is placed on a data store within the enterprise network. Unauthorized access to the honeypot file can then be detecting by monitoring use of the associated key material, which usefully facilitates detection of file access at any time when, and from any location where, cryptographic access to the file is initiated.

Abstract: In general, in one aspect, a method for machine learning recognition of portable executable files as malware includes providing training data comprising features of portable executable files and a descriptive information for the portable executable files, the descriptive information comprising a family or type of malware. The method may include training a model using the training data to detect malware. The method may include using the trained model to recognize malware by providing features of a portable executable file as input and providing a threat score and descriptive information as output.

Abstract: In some embodiments, a method includes processing at least a portion of a received file into a first set of fragments and analyzing each fragment from the first set of fragments using a machine learning model to identify within each fragment first information potentially relevant to whether the file is malicious. The method includes forming a second set of fragments by combining adjacent fragments from the first set of fragments and analyzing each fragment from the second set of fragments using the machine learning model to identify second information potentially relevant to whether the file is malicious. The method includes identifying the file as malicious based on the first information within at least one fragment from the first set of fragments and the second information within at least one fragment from the second set of fragments. The method includes performing a remedial action based on identifying the file as malicious.