Abstract: A cybersecurity engine can guide a forensic investigation of a security incident by estimating the utility of investigating events associated with the security incident, selecting a subset of such events based on the estimated utilities, and presenting data associated with the selected events to the investigator. A method for guiding a response to a security incident may include estimating, for each of a plurality of security events associated with the security incident, a utility of investigating the security event. The method may further include selecting a subset of the security events based, at least in part, on the estimated utilities of investigating the security events. The method may further include guiding the response to the security incident by presenting, to a user, data corresponding to the selected security events.

Abstract: The present system and method pertain to the detection of malicious software and processes such as malware. A cloud security policy system receives hashes and behavioral information about applications and/or processes executing on user devices. The cloud security policy system records this information and then evaluates the trustworthiness of the hashes based on the information received from the user devices to provide a security policy for the applications and/or processes. The security policy is sent from the cloud security policy system to user devices to be applied by the user devices.

Abstract: A security client can be configured to operate on the one or more computing systems and record all events occurring on the one or more computing systems. The security client can operate as a “security camera” for the computing systems by identifying and retaining data and information that describes and details different events that occur on the computing systems. The security client can be configured to generate event records for the events that are uniquely associated with the process that requested or performed event. Likewise, the security client can be configured to uniquely associate the event records with the specific computing system associated with the event.

Abstract: A system and method for tracking data security threats within an organization is proposed. A threat aggregator process executing on an analysis computer system within the organization receives events indicating possible threats observed by and sent from different user devices and aggregates related events into threats. This enables the threats to be analyzed and acted upon at a level of the organization (e.g., across user devices) rather than at the level of the individual user devices. An endpoint telemetry system analyzes threats sent from the aggregator and provides security policies for responding to the threats. In examples, the system can identify attacks of related threats and act upon the related threats of the attack collectively, and can characterize false positive threats sent from multiple user devices as a single extraneous threat. This has advantages over the per-user device focus for responding to threats provided by current systems and methods.

Abstract: Cybersecurity systems and techniques are described. A cybersecurity method may include generating a process fingerprint of a process, wherein the process fingerprint identifies the process based, at least in part, on dynamic features of the process. Generating the process fingerprint may include performing a cryptographic hash operation on data representing dynamic features of the process. The method may further include comparing the process fingerprint to a plurality of process fingerprints, and based, at least in part, on a result of the comparison, performing a data reduction operation on data associated with the process and/or determining whether the process is a malware process.

Abstract: A computing apparatus for protecting software hooks from interference may include a processing device and a memory access monitoring device configured to monitor access to the memory addresses of one or more hooks. When a task T1 attempts to write to a memory address of a monitored hook, the monitoring device may generate a notification (e.g., an interrupt), and the processing device may pause execution of the task T1 and initiate execution of a hook protection task T2. The hook protection task T2 may determine whether to allow task T1 to modify the monitored hook. If task T1 is not a trusted task (e.g., if task T1 is or may be malware), the processing device blocks T1 from modifying the monitored hook. In this manner, some attempts to unhook critical software hooks may be thwarted.

Abstract: A distributed security system and method are disclosed that enable access to known threat events from threat intelligence feeds when the system includes public cloud components. A cloud-based security policy system stores observable events for security incidents detected by and sent from user devices within an enterprise network. The observable events include observable indicators for characterizing the observable events. The threat events within the feeds include threat indicators for characterizing the threat events. An on-premises connector within the enterprise network downloads the observable indicators from the security policy system and the threat indicators from the feeds. In response to determining that any observable indicators match any threat indicators, the on-premises connector provides access to the threat events and/or the observable events having the matching indicators.

Abstract: A security engine may use event-stream processing and behavioral techniques to detect ransomware. The engine may detect process behavior associated with encrypting a file, encrypting a storage device, or disabling a backup file, and may assign a ransomware category to the process based thereon. The engine may initiate protection actions to protect system resources from the process, which may continue to execute. The engine may monitor the process for specific behavior corresponding to its ransomware category. Based on the extent to which such specific behavior is detected, the engine may determine that the process is not ransomware, assign a ransomware subcategory to the process, or adjust the process's threat score. Monitoring of the process may continue, and the threat score may be updated based on the process's behavior. If the threat score exceeds a threshold corresponding to the ransomware category (or subcategory), a corresponding policy action may be initiated.

Abstract: Behavioral baselines for a computer system may be accurately and efficiently established by (1) monitoring occurrences on the computer system, (2) determining, based on security rules or heuristics, which of the observed occurrences are associated with potential security risks, (3) identifying patterns of activity based on the suspicious occurrences, and (4) prompting a user to indicate whether the observed patterns of suspicious activity are expected or unexpected. Behavior baselines established in this manner can then be used to differentiate between expected and unexpected patterns of activity on the computer system.

Abstract: A cybersecurity engine can guide a forensic investigation of a security incident by estimating the utility of investigating events associated with the security incident, selecting a subset of such events based on the estimated utilities, and presenting data associated with the selected events to the investigator. A method for guiding a response to a security incident may include estimating, for each of a plurality of security events associated with the security incident, a utility of investigating the security event. The method may further include selecting a subset of the security events based, at least in part, on the estimated utilities of investigating the security events. The method may further include guiding the response to the security incident by presenting, to a user, data corresponding to the selected security events.

Abstract: A method and system for remediating a process hollowing intrusion on a user device comprising detecting a process starting on the user device, preparing the process to monitor Application Programming Interface (API) calls between the process and an operating system of the user device, determining whether the process is associated with a process hollowing intrusion based on information associated with the process and/or the API calls, and executing security policies against the process associated with the process hollowing intrusion. In examples, it is determined whether the child process is associated with a process hollowing intrusion in response to determining whether one or more API calls associated with known process hollowing intrusions modify executable memory of and/or modify an entry point address of the child process.

Abstract: A system and method for detecting reverse command shell intrusions at a process-level on a user device is disclosed. In one embodiment, the system detects each process starting on an operating system of the user device, such as a mobile phone or laptop computer, and monitors Application Programming Interface (API) calls between each process and the operating system. The system then determines whether each process is associated with a reverse command shell intrusion based on information associated with each process and/or the API calls, and executes security policies against the processes associated with the reverse command shell intrusion to remediate the processes. In another embodiment, the system determines whether processes starting on a user device are associated with a reverse command shell intrusion by monitoring and analyzing information associated with the parent process of each process and/or API calls between each parent process and the operating system.

Abstract: Cybersecurity systems and techniques are described. A cybersecurity method may include generating a process fingerprint of a process, wherein the process fingerprint identifies the process based, at least in part, on dynamic features of the process. Generating the process fingerprint may include performing a cryptographic hash operation on data representing dynamic features of the process. The method may further include comparing the process fingerprint to a plurality of process fingerprints, and based, at least in part, on a result of the comparison, performing a data reduction operation on data associated with the process and/or determining whether the process is a malware process.

Abstract: A system and method for tracking data security threats within an organization is proposed. A threat aggregator process executing on an analysis computer system within the organization receives events indicating possible threats observed by and sent from different user devices and aggregates related events into threats. This enables the threats to be analyzed and acted upon at a level of the organization (e.g., across user devices) rather than at the level of the individual user devices. An endpoint telemetry system analyzes threats sent from the aggregator and provides security policies for responding to the threats. In examples, the system can identify attacks of related threats and act upon the related threats of the attack collectively, and can characterize false positive threats sent from multiple user devices as a single extraneous threat. This has advantages over the per-user device focus for responding to threats provided by current systems and methods.

Abstract: A security client can be configured to operate on the one or more computing systems and record all events occurring on the one or more computing systems. The security client can operate as a “security camera” for the computing systems by identifying and retaining data and information that describes and details different events that occur on the computing systems. The security client can be configured to generate event records for the events that are uniquely associated with the process that requested or performed event. Likewise, the security client can be configured to uniquely associate the event records with the specific computing system associated with the event.

Abstract: The present system and method pertain to the detection of malicious software and processes such as malware. A cloud security policy system receives hashes and behavioral information about applications and/or processes executing on user devices. The cloud security policy system records this information and then evaluates the trustworthiness of the hashes based on the information received from the user devices to provide a security policy for the applications and/or processes. The security policy is sent from the cloud security policy system to user devices to be applied by the user devices.

Abstract: A cybersecurity engine can guide a forensic investigation of a security incident by estimating the utility of investigating events associated with the security incident, selecting a subset of such events based on the estimated utilities, and presenting data associated with the selected events to the investigator. A method for guiding a response to a security incident may include estimating, for each of a plurality of security events associated with the security incident, a utility of investigating the security event. The method may further include selecting a subset of the security events based, at least in part, on the estimated utilities of investigating the security events. The method may further include guiding the response to the security incident by presenting, to a user, data corresponding to the selected security events.

Abstract: A computing apparatus for protecting software hooks from interference may include a processing device and a memory access monitoring device configured to monitor access to the memory addresses of one or more hooks. When a task T1 attempts to write to a memory address of a monitored hook, the monitoring device may generate a notification (e.g., an interrupt), and the processing device may pause execution of the task T1 and initiate execution of a hook protection task T2. The hook protection task T2 may determine whether to allow task T1 to modify the monitored hook. If task T1 is not a trusted task (e.g., if task T1 is or may be malware), the processing device blocks T1 from modifying the monitored hook. In this manner, some attempts to unhook critical software hooks may be thwarted.

Abstract: Visual and non-visual elements associated with the candidate files are analyzed to determine whether the candidate files are malware. A visual element (e.g., icon) is extracted from the candidate file, and the icon's image is compared to a group of reference images associated with trusted entities. If the icon's image matches a reference image, the candidate file may be malware masquerading as trusted software. The non-visual elements associated with the candidate file are used, in combination with the visual elements, to determine whether the candidate file is malware.

Abstract: A security client can be configured to operate on the one or more computing systems and record all events occurring on the one or more computing systems. The security client can operate as a “security camera” for the computing systems by identifying and retaining data and information that describes and details different events that occur on the computing systems. The security client can be configured to generate event records for the events that are uniquely associated with the process that requested or performed event. Likewise, the security client can be configured to uniquely associate the event records with the specific computing system associated with the event.