System and method for automated system for triage of cybersecurity threats

Abstract

A device for verifying previous determinations from cybersecurity devices comprising a processor and a memory. The memory comprises submission analysis logic including workflow selector logic to receive the object data and process the object data to select at least one analyzer supported by the analyzer logic. The analyzer logic, in accordance with the selected analyzer(s), is configured to (i) analyze the object data for potential threats and embedded object data, (ii) generate results data based on that analysis, and (iii) pass the embedded object data back to the workflow selector for further analysis. Finally, the submission analysis logic comprises triage ticket generation logic to generate triage tickets for analyst review and alert logic to generate automatic alerts.

Description

FIELD

Embodiments of the disclosure relate to the field of cybersecurity. More specifically, one embodiment of the disclosure relates to a system, apparatus and method for automatically triaging submitted security-related data.

BACKGROUND

Over the last decade, malicious software (malware) has become a pervasive problem for Internet users. Often, malware is a program or file that is embedded within downloadable content and designed to adversely influence (e.g., attack) normal operations of a computer. Examples of different types of malware may include bots, computer viruses, worms, Trojan horses, spyware, adware, or any other programming that operates within the computer without permission.

For instance, malicious content may be embedded within objects associated with a web page hosted by a web site. By downloading and processing this malware (i.e. malicious content) another web page may be requested from a malicious web site and unknowingly installed on the computer. Similarly, malware may also be installed on a computer upon receipt, or opening, of an electronic mail (email) message. For example, an email message may contain an attachment, such as a Portable Document Format (PDF) document, with embedded executable malware. Also, malware may exist in files infected through any of a variety of attack vectors, which are uploaded from the infected computer onto a networked storage device such as a file share.

Over the past few years, various types of cybersecurity methods have been deployed that have attempted to find malware within files or other network content. Often, these methods evaluate objects suspected of being malicious and subsequently make a determination if the evaluated object is either a threat (i.e. malicious) or not a threat (i.e. benign). In certain circumstances, this evaluation process may yield false negative (a malicious object is deemed to be benign) or false positive (a benign object is deemed to be malicious) results. Such incorrect determinations can lead to users being otherwise exposed to potential threats.

In order to minimize such false positives and negatives, many systems can employ a team of human analysts to further evaluate objects for potential threats. In certain instances, users and/or system administrators of security system products submit an object to a security system vendor to ask for an evaluation or re-evaluation. These submissions may be items that are known to be benign that were deemed to be malicious by the security system during analysis or is perhaps a suspect item that was deemed benign by the security system. In these cases, the submissions are evaluated by the human analysts with a variety of tools to further understand how and/or why the security system came to the conclusion it did.

Such supplemental analysis can be time-consuming and an inefficient use of the human analysts' time as a multitude of analyses can occur simultaneously. In many cases, the human analyst is given the suspicious item without any further background information beyond what the customer provided in the submission process. Often, the lack of data provided by the user in the submission process creates further inefficiencies in the human analysts' time.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention are illustrated by way of example and not by way of limitation in the figures of the accompanying drawings, in which like references indicate similar elements and in which:

FIG. 1 is an exemplary system diagram of a communication network with a plurality of cybersecurity and network devices in communication with a submission analysis system device.

FIG. 2A is an exemplary hardware block diagram of a cybersecurity device.

FIG. 2B is an exemplary hardware block diagram of a submission analysis system device.

FIG. 3 is an exemplary flowchart of a triage data generating process.

FIG. 4 is an exemplary flowchart illustrating the interoperation between the submission analysis logics.

FIG. 5 is an exemplary embodiment of a flowchart illustrating the generation of triage data in response to a potential false negative and/or false positive result.

DETAILED DESCRIPTION

In light of the problems discussed above, a solution is needed to accelerate the supplemental analysis and classification of objects suspected of being falsely classified. Various embodiments of the disclosure relate to a system and/or process configured to analyze submitted objects and triage the analyzed objects. The triage process can occur by performing analytics on a submitted object, correlating any results, assigning a score and providing that information to an alert logic. The triaged objects can also be utilized to generate triage tickets which may aid human analysts by providing a preliminary set of data to aid in their analysis. Furthermore, the triage system and/or process may comprise a set of pre-determined thresholds or rules that can allow for automatic responses to be generated when presented with certain results without the need for intervention by a human analyst.

More generally, the cybersecurity triage system described herein can receive messages comprising object data and context data associated with classification of the object. The object data and context data are received from a submitter (e.g., a user, customer, etc.) and parsed for re-evaluation of a potential cybersecurity threat. The context data may include meta information about the submitter cybersecurity threat detection system as well as a classification (i.e. verdict) associated with the object. The cybersecurity triage system may reanalyze the received object data in a plurality of analyzers, the analyzers selected by a workflow selector based on at least the object and context data. Results from the selected analyzers are correlated and scored for cybersecurity threats based at least in part on stored, prior classifications. Consequently, triage data is generated. Based on the triage data, a response logic may determine if an alert can automatically be generated to provide (1) the submitter with confirmation of a submitted classification, (i.e. the submitted classification was correct and not a false positive or false negative) (2) the submitter with a correction of submitted classification (i.e. the submitted classification was incorrect and was thus a false positive or false negative), or (3) a cybersecurity analyst with at least the triage data and object data. The cybersecurity triage system thereby, provides an automated triage system based on reevaluation of the received object data to accelerate and improve the efficiency of cybersecurity threat detection support systems.

The cybersecurity triage system can exist within a larger network that may comprise a plurality of network devices having access to a cybersecurity system which may, among other things, analyze incoming objects to determine if a threat is posed to the cybersecurity system or network devices.

Often, the threat classification is based on the security content present in the customer's computing environment. The security content may include, but is not limited or restricted to, analyzer data, scanning rules, heuristics, other digital signatures indicative of threats or any combination thereof. In certain embodiments, the original classification may have been determined by a first cybersecurity system and the reevaluation is subsequently performed by a second and different cybersecurity system or at least by a cybersecurity system with an alternate configuration. In this way, the submission analysis system may attempt to verify the results of prior classifications for customers who are suspicious about the accuracy of such prior threat classifications.

In cases where the classification of a threat is considered to be incorrect by a cybersecurity system user (which may include general users or system administrators), an alert system may allow the cybersecurity system user to generate a submission message that may contain the suspect object (or a hash associated with the suspect object) and send it to a submission analysis system operated by a cybersecurity system vendor (such as, for example, a managed services provider) and/or manufacturer who may then reevaluate the suspicious object to determine if the cybersecurity system generated a false positive or false negative.

The submission message utilized for sending the suspicious object to the submission analysis system for a subsequent classification can be generated on a cybersecurity system configured to create the submission message. The message submission process may be realized through a variety of means including, but not limited to, specialized software agents running on a host computing device, web-based applications hosted on a local cybersecurity system, web-based applications hosted on a remote submission analysis system, and/or hardware interfaces or appliances located within the network. It should also be apparent to those skilled in the art that submission messages may be generated manually or through an automated process (typically in response to at least one pre-determined rule). Additionally, further data relating to the system context can be generated relating to the current operating environment (computer model/operating system/system settings/network settings/etc.) and software related data (file type/associated software/software version/software settings/etc.). Such context data can be packaged with the submission message.

The submission analysis system can receive and parse the incoming submission message data in a manner that allows the system to evaluate and manipulate the suspected object embedded within the submission message data. In many embodiments, this manipulation includes generating and/or instantiating a number of (e.g., one or more) analyzers that can process the object data to evaluate threats and/or determine the methods by which a threat determination was made on the system which submitted the data. The various system and software settings available within the context data can be utilized during evaluation to aid the analyzer logic which may then select and generate analyzers that can mimic the context data settings from the system submitting the message in order to analyze the submitted object similarly to how it may have been processed within the submitting system.

Typically, analyzers are generated by the system to perform analytics on a suspicious object, which may comprise inspecting, identifying and analysis. Such analysis may further comprise examining behaviors of the object that may occur through processing the suspicious object within a virtualized environment and/or through an emulation process, comparing the hash of the suspicious object with hashes of other known malicious and/or benign objects, examining the content of the object to explore its code and structure, as well as looking at any context or metadata that may provide further insight into the suspicious object.

Furthermore, certain embodiments may generate additional analyzers to evaluate the same object data under different computing and/or software environments to determine if similar findings would have occurred. By way of example and not limitation, an object that underwent analytics operating within a first version of a software package may yield a different threat determination than the same object analyzed operating within a second version of the same software. In this way, the submission analysis system may compare threat analyses of objects under a number of different system and software settings. This evaluation under different environments and settings may yield significant insights into the causes of false positive and/or false negative determinations, which can be utilized to generate triage data usable by a human analyst in further evaluations or by the submission analysis system to undertake an automatic remedial action without the need for intervention by a human analyst.

One practical feature of the invention is to automatically generate triage data that can be utilized by human analysts. In this way, the use of the submission analysis system can save the time needed by the human analyst when evaluating the submitted object. Another practical feature is to allow the cybersecurity triage system to process the triage data and automatically generate reports or undertake remedial action based on at least one pre-determined rule without the need for human analysts By having triaging data generated automatically prior to intervention by a human analyst, potential solutions can be provided to the submitting user with a shortened turnaround time, avoiding further potential threats that were deemed benign and allowing the human analyst to focus on other analyses that cannot otherwise be automatically determined by the system.

I. Terminology

In the following description, certain terminology is used to describe features of the invention. For example, in certain situations, the terms “logic” and “engine” are representative of hardware, firmware or software that is configured to perform one or more functions. As hardware, logic may include circuitry such as processing circuitry (e.g., a microprocessor, one or more processor cores, a programmable gate array, a microcontroller, an application specific integrated circuit, etc.), wireless receiver, transmitter and/or transceiver circuitry, semiconductor memory, combinatorial logic, or other types of electronic components.

As software, logic may be in the form of one or more software modules, such as executable code in the form of an executable application, an application programming interface (API), a subroutine, a function, a procedure, an applet, a servlet, a routine, source code, object code, a shared library/dynamic load library, or one or more instructions. These software modules may be stored in any type of a suitable non-transitory storage medium, or transitory storage medium (e.g., electrical, optical, acoustical or other form of propagated signals such as carrier waves, infrared signals, or digital signals). Examples of non-transitory storage medium may include, but is not limited or restricted to a programmable circuit; semiconductor memory; non-persistent storage such as volatile memory (e.g., any type of random access memory “RAM”); persistent storage such as non-volatile memory (e.g., read-only memory “ROM”, power-backed RAM, flash memory, phase-change memory, etc.), a solid-state drive, hard disk drive, an optical disc drive, or a portable memory device. As firmware, the executable code is stored in persistent storage.

The term “network content” generally refers to information transmitted over a network as one or more messages, namely a grouping of information that comprises a header and a payload, such as any of the following: a packet; a frame; a stream being a sequence of packets or frames; or any other series of bits having a prescribed format. The “payload” is generally defined as including the data associated with the message such as text, software, an image, an object, audio, video, a Uniform Resource Locator (URL), or other types of digital data. The “header” is generally defined as including control information. However, the specific types of control information depend on the network content type.

Other examples of network content include web traffic Hypertext Transfer Protocol (HTTP), email, which may be transmitted using an email protocol such as Simple Mail Transfer Protocol (SMTP), Post Office Protocol version 3 (POP3), or Internet Message Access Protocol (IMAP4). A further example of network content includes an Instant Message, which may be transmitted using Session Initiation Protocol (SIP) or Extensible Messaging and Presence Protocol (XMPP) for example. Yet another example of network content includes one or more files that are transferred using a data transfer protocol such as File Transfer Protocol (FTP) for subsequent storage on a file share. Where the network content is email, Instant Message or a file, the header may include the sender/recipient address, the sender/recipient phone number, or a targeted network location of the file, respectively.

The term “malware” is directed to software that produces an undesirable behavior upon execution, where the behavior is deemed to be “undesirable” based on customer-specific rules, manufacturer-based rules, or any other type of rules formulated by public opinion or a particular governmental or commercial entity. This undesired behavior may include a communication-based anomaly or an execution-based anomaly that (1) alters the functionality of an electronic device executing that application software in a malicious manner; (2) alters the functionality of an electronic device executing that application software without any malicious intent; and/or (3) provides an unwanted functionality which is generally acceptable in other context.

The term “object” generally refers to content in the form of an item of information having a logical structure or organization that enables it to be classified for purposes of analysis for malware. One example of the object may include an email message or a portion of the email message. Another example of the object may include a storage file or a document such as a Portable Document Format (PDF) document, a word processing document such as Word® document, or other information that may be subjected to cybersecurity analysis. The object may also include an executable such as an application, program, code segment, a script, dynamic link library “dll,” URL link, or any other element having a format that can be directly executed or interpreted by logic within the electronic device. Network content may be an example of an object and used for malware detection.

Logic (or component or module) may be software that includes code being one or more instructions, commands or other data structures that, when processed (e.g., executed) to perform a particular operation or a series of operations. Examples of software include an application, a process, an instance, Application Programming Interface (API), subroutine, plug-in, function, applet, servlet, routine, source code, object code, shared library/dynamic link library (dll), or a collection of HTML elements. This software may be stored in any type of a suitable non-transitory storage medium, or transitory storage medium (e.g., electrical, optical, acoustical or other form of propagated signals such as carrier waves, infrared signals, or digital signals). Examples of non-transitory storage medium may include, but are not limited or restricted to a programmable circuit; non-persistent storage such as volatile memory (e.g., any type of random access memory “RAM”); or persistent storage such as non-volatile memory (e.g., read-only memory “ROM”, power-backed RAM, flash memory, phase-change memory, etc.), a solid-state drive, hard disk drive, an optical disc drive, or a portable memory device. As firmware, the logic (or engine/component) may be stored in persistent storage.

The term “transmission medium” is a communication path between two or more systems (e.g. any electronic devices with data processing functionality such as, for example, a security appliance, server, mainframe, computer, netbook, tablet, smart phone, router, switch, bridge or router). The communication path may include wired and/or wireless segments. Examples of wired and/or wireless segments include electrical wiring, optical fiber, cable, bus trace, or a wireless channel using infrared, radio frequency (RF), or any other wired/wireless signaling mechanism.

The term “network device” should be generally construed as electronics with data processing capability and/or a capability of connecting to any type of network, such as a public network (e.g., Internet), a private network (e.g., a wireless data telecommunication network, a local area network “LAN”, etc.), or a combination of networks. Examples of a network device may include, but are not limited or restricted to, the following: a server, a mainframe, a cybersecurity device, a firewall, a router; an info-entertainment device, industrial controllers, vehicles, or an endpoint device (e.g., a laptop, a smartphone, a tablet, a desktop computer, a netbook, gaming console, a medical device, or any general-purpose or special-purpose, user-controlled electronic device). In many embodiments, network devices may contain cybersecurity logic that can generate submission messages.

The term “message” generally refers to signaling (wired or wireless) as either information placed in a prescribed format and transmitted in accordance with a suitable delivery protocol or information made accessible through a logical data structure such as an API. Examples of the delivery protocol include but are not limited or restricted to HTTP (Hypertext Transfer Protocol); HTTPS (HTTP Secure); Simple Mail Transfer Protocol (SMTP); File Transfer Protocol (FTP); iMessage; Instant Message Access Protocol (IMAP); or the like. Hence, each message may be in the form of one or more packets, frame, or any other series of bits having the prescribed, structured format.

Lastly, the terms “or” and “and/or” as used herein are to be interpreted as inclusive or meaning any one or any combination. Therefore, “A, B or C” or “A, B and/or C” mean “any of the following: A; B; C; A and B; A and C; B and C; A, B and C.” An exception to this definition will occur only when a combination of elements, functions, steps or acts are in some way inherently mutually exclusive.

As this invention is susceptible to embodiments of many different forms, it is intended that the present disclosure is to be considered as an example of the principles of the invention and not intended to limit the invention to the specific embodiments shown and described.

II. General Architecture

Referring to FIG. 1, an exemplary system diagram of a communication network 100 with a plurality of cybersecurity and network devices 110₁, 110₂, 110_N, 115 in communication with a submission analysis system device 130 is shown. The network 100 may include a first cybersecurity device 110₁ connected to the network 120 that may act as a stand-alone device or may also be commutatively coupled with a second cybersecurity device 110₂. It is contemplated and implied in FIG. 1 that any number of cybersecurity and network devices 110₁-110_N, 115 may be commutatively coupled to the network 120 and thus to the submission analysis system 130. In many embodiments, the cybersecurity and network devices 110₁-110_N, 115 can comprise cybersecurity logic 116 that can evaluate suspicious objects and create determinations on their potential threat levels.

Cybersecurity and network devices 110₁-110_N, 115 may communicate over private and/or public transmission mediums through the network 120 including the Internet, direct connections between the devices and/or via a wireless network access points. The cybersecurity devices 110₁-110_N and/or network devices 115 may be various computing devices such as, but not limited to, desktop computing systems, enterprise computing systems, electronic portable computers, or portable cellular phones.

Each submission analysis system device 130 can be adapted in many embodiments to receive submission messages and process the embedded code samples taken from network content for potential malicious code in real-time or at least near real-time. In certain embodiments, the submission messages can be provided in an offline process such as, but not limited to, a user presenting the data to be processed in the form of a drive coupled to the submission analysis system 130.

According to the embodiment depicted in FIG. 1, each cybersecurity device 110₁-110_N and network device 115 may contain and/or be in communication with a security appliance or agent that is configured to inspect ingress data traffic, identify whether any objects of the data traffic may include malware, and if so, analyze at least those objects. In certain embodiments, this analysis may be conducted in a virtual machine, but may be conducted without the aid of a virtual machine execution environment to detect anomalous behaviors that would be present if the data traffic was actually processed by an electronic device. As a result of the inspections done by the security appliances, objects will be determined to either be malicious or benign.

There are numerous methods in which a security appliance may classify objects as being malicious or benign. In some instances, determinations may depend on various settings within system software. At times, determinations may yield “malicious” or “benign” results. A user operated network device, security administrator, and/or an automated system relying on heuristics or thresholds of confidence in the classification system may suspect the classification is inaccurate (e.g. if the first classification was “malicious” however, the object is suspected of being benign, the object may be considered a “false positive”). This can lead to situations in which an end user is alerted to an object or network traffic being flagged as malicious, even when it is known that the flagged item is in fact benign. Alternatively, a user may still feel unsure of a suspicious item that was determined to be benign. In such situations a submission of these items to an external party such as, but not limited to, the security applicant vendor or manufacturer for reevaluation may be generated. Such submissions can be packaged via a submission message software and sent to a submission analysis system device 130 for evaluation prior to a human analyst looking at the submission.

The submission analysis system 130 can be comprised of a variety of logics and data stores that can be utilized to generate triage data based on received submission messages. In the embodiment depicted in FIG. 1, the submission analysis system 130 is comprised of workflow selector logic 145, object parsing logic 140, analyzer logic 150, correlation engine logic 155, triage ticket generation logic 160 along with a storage device 170, and alert logic 180. As those skilled in the art will recognize, the specific naming, order, and number of logics may vary within the submission analysis system 130 based on the desired application.

It is contemplated that the submission analysis system 130 may be deployed as a device but may also be implemented in a cloud computing service for analysis of the submitted data as described. Furthermore, it is contemplated that the functionality of one or more submission analysis system devices 130 may be incorporated into another management system when malware detection is to be conducted at a centralized resource (e.g. in a load balancing system). By way of example and not limitation, a private organization may incorporate a submission analysis system 130 to re-evaluate threat determinations done on its own systems in order to generate more threat analysis data, keep certain submission objects internal (e.g., where data privacy is an issue), and/or to reduce the time required by human analysts to evaluate the results by avoiding reliance on a third party for the reevaluation.

Referring now to FIG. 2A, an exemplary hardware block diagram of a cybersecurity device 110₁ is shown. The cybersecurity device 110₁ and respective hardware can be similar to the network devices 110₂, 110_N as depicted and discussed in FIG. 1. The cybersecurity device 110₁ comprises one or more processors (e.g., CPU) 210, which is coupled to communication input/output interfaces 230 as well as to a persistent memory system 220. Within the persistent memory system 220 of the embodiment depicted in FIG. 2A, various logics may be provided including object submission logic 205 and context data logic 206. It is also contemplated that certain embodiments of the cybersecurity logic 116 operating within a network device 115 may also perform functions similar to those described with reference to the network device 110₁ illustrated in FIG. 2A.

In additional embodiments, the object submission logic 205 can be utilized by the user to package, generate, and send a submission message containing objects for reevaluation by a submission analysis system. This process may be effected through the use of a user interface that prompts the user for information related to the object and reasons for submission. In certain embodiments, the object submission logic 205 may automatically generate the submission with data based on the initial evaluation by the cybersecurity system. In these instances, the system may automatically send the submission or could prompt the user to verify the automatically generated data prior to sending the submission. Additionally, the choice between manual and automatic submission message generation and sending may be determined through the use of a pre-determined rule which may include, but is not limited to, the type of threat, the number of evaluations done over a given period of time, time of day generation occurs, and/or the security level of the given user.

Context data logic 206 can be utilized to further add supplemental data to the submission message related to the state of the submitting computing system. In various embodiments, the added context data may aid in the determination of why certain evaluations occurred within the cybersecurity system. Context data may include a variety of computing environment data including, but not limited to, operating system data, operating system settings, network settings, network security settings, local evaluation data, software data, software settings, and/or software version. It would be obvious to those skilled in the art that context data may include other forms of data not specifically outlined above as long as said data may aid in the reevaluation of the object. In some embodiments, context data may include data derived from an external source including, but not limited to, a global data store system similar to those described in U.S. patent application Ser. No. 16/222,194 filed Dec. 17, 2018 and entitled “Platform and Method for Enhanced Cyber-Attack Detection and Response Employing a Global Data Store” the disclosure of which is hereby incorporated by reference in its entirety.

Referring now to FIG. 2B, an exemplary hardware block diagram of submission analysis system device 130 is shown. The submission analysis system device 130 comprises one or more processors (e.g., CPU) 250, which are coupled to communication input/output interfaces 270 as well as to a persistent memory system 260. Within the persistent memory system 260 of the embodiment depicted in FIG. 2, various logics may be provided including submission analysis logic 240. The submission analysis logic 240 further comprises object parsing logic 140, workflow selector logic 145, analyzer logic 150, correlation engine logic 155, triage ticket generation logic 160, and alert logic 180.

The input/output interface 270 can enable communications with other components including other cybersecurity and network devices 110₁-110_N, 115 as depicted in of FIG. 1 as well as other electronic devices over private and/or public networks. According to one embodiment of the disclosure, the input/output (I/O) interface 270 may be implemented as a physical interface including one or more ports for wired connectors. Alternatively, the entire submission analysis system device 130 can be implemented as software and the input/output interface is implemented as a digital communication logic between the submission analysis system device 130 and other software interfaces. Additionally, or in the alternative, the I/O interface 270 may be implemented with one or more radio units for supporting wireless communications with other electronic devices.

In certain embodiments, the processor 250 can be communicatively coupled to persistent storage 260 over an external transmission medium via the I/O interface 270. According to one embodiment of the disclosure, the persistent storage 260 may load the submission analysis logic 240 within a volatile memory for execution. Of course, in embodiments where the logic is implemented as hardware, submission analysis logic 240, including the sub-components would be implemented separately from persistent memory 260.

The submission analysis logic 240 may be comprised of many sub-components or sub-systems with internal logic suited for particular tasks. In one embodiment illustrated in FIG. 2B, the submission analysis logic 240 comprises separate object parsing logic 140, workflow selector logic 145, analyzer logic 150, correlation engine logic 155, triage ticket generation logic 160, and alert logic 180.

In many embodiments, the object parsing logic 140 can accept submission messages and parse them for processing and evaluation. The parsing process can identify both the object that is to be evaluated and any context and/or other meta information that may aid in the evaluation of the submission message. By way of example and not limitation, the object parsing logic 140 may accept a submission message comprising object data and context data. The object parsing logic 140 may then determine, for example, that the object data comprises a portable document format (“PDF”) file as the object to be evaluated as well as context data that was generated by the agent on the cybersecurity device 110 or network device 115 which informs the submission analysis device about the computing environment the PDF was obtained from. It would be known to those skilled in the art that numerous types of objects may be parsed from the submission message.

The software agent utilized for the generation of the submission messages can vary based on the type of device it operates. In a number of embodiments, the submission message agent may operate automatically on an intermediary network device. In other embodiments, the submission message agent may allow for more direct feedback and formatting from a user when located on an endpoint network device. It would be apparent to those skilled in the art that such software agents can be configured to a variety of configurations based upon the application desired and may even be deployed as a single software package that can adapt based on the type of device and/or computing environment it operates within.

Workflow selector logic 145, in a variety of embodiments, can receive and analyze object and context data to select analyzers for evaluation. In certain embodiments, the workflow selector logic 145 may examine the object data for an object data type and select a first sub-set of analyzers for use in the analyzer logic 150. Furthermore, the workflow selector logic 145 may also examine the context data to select a second sub-set of analyzers from within the selection of analyzer logic 150. By way of example and not limitation, the workflow selector logic 145 can analyze received object data and determine that the object is an email type, which yields a first selection of analyzers from within the analyzer logic 150. In many embodiments, the workflow selector logic 145 may implement an ordered analysis of the object. In various embodiments, the analysis may be done in parallel among a plurality of analyzers with their results able to be correlated by the correlation engine logic 155. In response, the workflow selector logic 145 may then analyze the context data and determine what version of email client the cybersecurity device was using or evaluating. In response to this, the workflow selector logic 145 may select a second set of analyzers from within the analyzer logic 150 that correspond to different versions of the email client. Selected analyzers may analyze the object within multiple environments wherein the software versions are different. Through this process, the differences between the processing of the object through different versions may yield data related to why the cybersecurity device came to the threat determination it previously did. In this way, the submission analysis system can utilize the differences within the data generated from the analyzers to generate triaging data which may be utilized to make further determinations regarding the validity of the first determination.

Similarly, it is contemplated that other types of selections may be made based on the contents of the object and/or context data. These other selections may be in response to a variety of factors including, but not limited to, application settings, operating system, operating system settings, file source, and/or file destination. Selection data can be generated by the workflow selector logic 145, which may then be passed to the analyzer logic 150 for implementation. In some embodiments, selection data may comprise data indicating particular analyzers that can be utilized by the analyzer logic 150.

In a number of embodiments, the analyzer logic 150 includes a plurality of different analyzers that may be utilized to evaluate object data. The analyzer logic 150 can receive selection data from the workflow selector logic 145 and launch or utilize the selected analyzers. In some embodiments, the analyzer logic 150 may utilize pre-generated analyzers available in other systems. In certain embodiments, the analyzers may be behavioral analyzers comprising instrumented virtual machines to process the object. The analyzers can be configured according to the context data on which to base the settings of the analyzer. In additional embodiments, the analyzers may emulate the computing environment selected in the selection data and monitor various settings such as, but not limited to, buffer, memory reads/writes, and/or function calls and returns. It is contemplated that the analyzers utilized by the analyzer logic may include a variety of analyzers and may also include other analyzer types based upon the application needed.

The composition or type of analyzers may be updated or otherwise changed periodically or aperiodically to allow for improved analyzers, new analyzers, or removal of ineffective analyzers. The updates may be remotely pushed or can be locally updated through a network connection or a physical connection with an update drive. The output of the analyzers can be passed to the correlation engine logic 155 in the form of results data. Results data may comprise descriptions of any changed states, variables, or behaviors within the analyzed objects.

In a variety of embodiments, correlation engine logic 155 examines and scores the results data from the analyzers into a set of score data. In further embodiments, the score data may be generated by examining the differences between the various analyzers. By way of example and not limitation, the results data can be generated from a variety of analyzers that were instantiated to emulate or simulate a number of different configuration settings within the same email client program used by the cybersecurity device 110₁ that flagged the email as a threat. Examining the results of the various analyzers may indicate the many analyzers had come to the same conclusion as the cybersecurity device 110₁; however, it may also indicate that certain analyzers with particular configuration settings changed determined the email was not a threat. In this way, the score data may be formatted and adapted to reflect that various differences in the ultimate processing results in the set of analyzers, which data can be further processed to yield additional insights, such as data indicating accuracy of the various analyzers.

Triage ticket generation logic 160, in certain embodiments, can accept the score data and determine if a triage ticket should be generated. A set of pre-defined thresholds relating to the circumstances warranting a triage ticket to be generated can be pre-loaded into the submission analysis system 130. It would be known to those skilled in the art that the pre-defined threshold may, in additional embodiments, be variable or otherwise updatable through a variety of means including updated rule sets, machine learning, and/or neural networks that can update the weights or rules based on a given input data set. By settings these thresholds and rules, the triage ticket generation logic 160 can evaluate when insightful data was gained from the analyzers and can generate and format a triage ticket reflecting the insightful data which can be reported to a human analyst for further analysis. In this way, the generation of a triage ticket can alert the human analyzer to focus on a specific area of concern based upon the results data generated by the analyzers, thus saving time and effort.

There may be instances where a remedial action may need to be taken before a human analyzer has time to read and analyze a triage ticket. In these embodiments, the submission analysis logic 240 may comprise an alert logic 180 that can take remedial actions without human intervention. In many embodiments, the submission analysis device 130 may be given a set of pre-defined thresholds and/or rules that may empower the submission analysis device 130 to take remedial actions immediately based on the score data derived from the results data. By way of example and not limitation, the correlation engine logic 155 may generate score data that indicates that a simple change in the settings of the computing environment or software suite of the cybersecurity device 110₁ can yield a satisfactory result for the user. The alert logic 180 can then auto-generate a report, email, or other communication to the user informing them of these findings and providing a suggestion to change the setting that was causing the negative outcome. In other cases, the alert logic 180 may respond to the determination that the score data indicates that a false negative occurred and that it would be advantageous to notify all potentially affected parties as soon as possible in case a human analyst was not immediately available to evaluate. Thus, the alert logic could report out to not only users, but also to outside or third parties that may benefit from this information.

III. Basic Triage Data Generation

Referring to FIG. 3, an exemplary flowchart of a triage ticket generating process is shown. The process 300 as outlined can be accomplished by a variety of devices and methods, including by the devices outlined in FIGS. 2A-2B. The process 300 typically begins when a user generates a submission message for reevaluation (block 310). Once generated, the submission message can then be transmitted to the submission analysis system for processing (block 320). Upon reception, the submission analysis system can then parse the message to extract the object that is to be reevaluated along with any other relevant data packaged within the submission message (block 330). In many embodiments, the submission message is parsed into the object data and context data that comprises data related to the host computing and/or software environment.

The submission analysis system can take the submitted data, once parsed, and analyze it to select a workflow that may best process the data (block 340). In a number of embodiments, the workflow selection is a selection of particular analyzers within the analyzer logic. In further embodiments, the selection may examine the object to determine the file type or contents of the object data and select an analyzer that has been designed to best analyze that file type or contents. By way of example and not limitation, the selection may be based on the determination that the object data for reevaluation contains a text file, which then selects an analyzer that is best suited to analyze text files.

Once selected, the process 300 may then process the object with the selected analyzers (block 350). In certain embodiments, the analyzers may include a virtual machine that attempts to open, process, or otherwise operate the object by common (or otherwise expected) means. In additional embodiments, the analyzers may analyze the object data in an emulated environment. In further additional embodiments, the analyzers may generate a hash of the object data for further comparison/analysis.

In certain instances, the object data, upon analysis in the analyzer, may generate or locate an embedded object that itself may need to be evaluated. By way of example and not limitation, a text file may contain a link to an external website, which itself should be evaluated as part of the overall threat analysis. As such, the process 300 determines if new objects have been found in process within the analyzed object (block 360). When embedded objects are found, the process 300 can send the newly determined object data to the workflow selector to then again select a workflow based upon the embedded object (block 340). When all objects have been found and no further embedded objects are determined, the process 300 may then correlate the results and generate a score (block 370). Such correlation may include examining the analyzed results with results previously provided in the context data of the submitting machine. The correlation of the results can be quantified as a series of score data. Upon completion of the results correlation, the process 300 may classify the object based on the score which can then be utilized to make a reevaluation (block 380). The classification can be realized through a variety of means including comparison to a set of pre-determined thresholds or rules

In many embodiments, the process 300 determines if the triage ticket should be generated based upon the classification of the score data (block 385). In certain cases, the score may be classified in a way that does not warrant the generation of a triage ticket. In more embodiments, score data can be utilized to generate a triage ticket (block 390). The exact type of triage ticket generated can depend on a plurality of pre-determined rules that may interpret and/or classify the score data. In response to the score being classified such that a triage ticket does not need to be generated (and consequently not analyzed by a human analyst), an automatic response may be generated and sent back to the submitting user (block 395). By way of illustrative example, the results may yield a determination that the suspicious object is benign and generated a false positive result due to differences in how the submitter's software version handles the object. In these cases, no direct involvement of the human analyst is needed and an automatic response with remedial actions to take may be generated and sent to the submitter. In this way, the human analyst may increase the amount of time spent on other undetermined threat analyses. The flow of the various logics of the submission analysis system is described in more detail in FIG. 4.

IV. Submission Message Data Flow

Referring now to FIG. 4, an exemplary logic flowchart detailing the processes of processing submission message data with submission analysis system logic is shown. In many embodiments, a submission message 410 is generated that comprises both object data 411 representing the object under reevaluation, and context data 412 which can itself comprise data relating to the submitting computing device environment and/or settings.

Upon submission through any of a variety of transmission mediums, the parsing logic 420 can then parse and process the submission message into its constituent object data and context data parts in a format that can be utilized within the submission analysis logic. Parsed data is then sent to workflow selection logic 430 that can then determine the most suitable analyzer for the object data. In certain cases, the object data may have multiple objects that require multiple analyzers of different types.

The object data is passed to the analyzer logic 440 which may then generate any number of selected analyzers A-N 441, 442, 443, 444 based on the type of object data submitted. As discussed above, object data analyzed within an analyzer may have another embedded object that must also have an analyzer selected. In these instances, the embedded object is sent back to the workflow selector logic for further evaluation and analyzer selection. In other instances, object data may be evaluated through several analyzers to determine differences in results.

By way of example and not limitation, object data may be emulated or run within a virtual environment or with a first version of a particular software application, along with a second or alternative version, etc. In some embodiments, the analyzer logic 440 may model a computing system to generate results data based on observed behaviors or instructions associated with the suspicious object. In certain embodiments, the analyzer logic 440 generates virtualized machines that simulate a run-time environment expected by the suspicious object in order to perform in-depth dynamic (behavioral) analysis on the object in efforts to detect one or more anomalous behaviors. The virtual machines may be configured with a software or computing environment profile corresponding to an image that can be configured or selected by workflow selector logic 430. The analyzers evaluate the processing of the object data and generate results data.

The results data can be passed through to correlation engine logic 460 that may then analyze the entire set of data between the plurality of analyzers 441, 442, 443, 444. The differences between the various analyzers 441, 442, 443, 444 may be utilized to generate score data. In many embodiments, score data can additionally be generated by evaluating the parsed context data 412 within the submission message 410. In some embodiments, the correlation engine logic 460 may be configured to receive an external data feed 450, which can be utilized to add further information (e.g. context) during correlation and added as part of the score data. By way of example and not limitation, the external data feed 450 may be comprised of newly discovered threats and/or issues related to updated computing/software environments and settings subsequently identified by any source including a third-party source. Similarly, in some embodiments, classifications associated with previously analyzed objects may be used by the correlation engine logic 460 to add further context to the score data.

Upon generation of score data, it can be utilized by the triage ticket generation logic 470 which can evaluate the score data to determine if a triage ticket should be generated, or some other method of reporting the results to a human analyst. In a number of embodiments, triage data generated may be generated and utilized to create a triage ticket as part of an auto-generated report that is readable by the human analyst. In response to at least one pre-determined rule, triage data may be utilized by alert logic 480 to auto-generate an alert that can be sent out to the human analyst for immediate review, to the user of the submitting device, and/or to third parties.

V. Detailed Triage Ticket Data Generation Process

Referring now to FIG. 5, an exemplary embodiment of a flowchart illustrating the generation of triage data in response to an object re-evaluation such as a potential false negative and/or false positive result is shown. The process can begin when a user receives a notification from their cybersecurity system or network device containing cybersecurity logic that a false negative or false positive result was found (block 500). In response, the user may generate a submission message to analyze the potential false negative and false positive result (block 505). In some embodiments, the user can generate the submission message manually through the use of a software agent specialized to generate submission messages for the submission analysis system. In additional embodiments, the software agent may automatically evaluate the host system and generate the submission message automatically based on a set of pre-determined thresholds and/or rules.

The submission message, prior to sending, can be formatted with object data reflecting the object that is to be reevaluated and context data reflecting the computing environment of the cybersecurity device (block 510). It is contemplated that the submission message format may comprise additional types of data based upon the specific application needed. The submission message, once generated, is transmitted to the submission analysis system where it is parsed for further processing (block 515).

In a number of embodiments, the submission analysis system may make a pre-analysis determination if the submission message is satisfactory, eligible, and/or warrants a full submission analysis (block 520). This step may allow the submission analysis system to reject any messages that are not formatted correctly, or that have already previously been evaluated. It is contemplated that the submission analysis system can have access to a data store comprising data that relates to past message submissions and eventual results. Once the submission analysis system determines that the submission message should be fully processed, the parsed object and context data can be sent to the workflow selector logic.

The workflow selector logic accepts the object and context data and evaluates them to select at least one analyzer from the selection of analyzers available in the analyzer logic (block 525). The selection is typically done based on a combination of object and context data contents. In certain embodiments, the selection may be composed as an intersection of the object and context data contents, whereas other embodiments may determine an additive selection process may yield better potential results.

Once selected, the selection data is sent to the analyzer logic which in response generates (or instantiates) the selected analyzers (block 530). The analyzers can mimic certain computing environments that may be similar but are usually slightly different from the computing environment of the cybersecurity device that generated the submission message. As discussed above, analyzers can emulate various computing environments, operating systems, software applications, along with a variety of operating system/software settings and varying versions. Typically, multiple analyzers are utilized to parse a single object so that differences in the output or operation can be observed.

In certain instances, when parsing the object data, an embedded object is discovered that may link or contain another object that should be evaluated (block 535). In these instances, the process can send the newly discovered object back to the workflow selector logic for analysis to determine further analyzers to process the embedded object. In some embodiments, new context data may be extracted along with the embedded object that may be passed to the workflow selector logic. In the absence of additional context data specific to the embedded object, the parent object's context data may be utilized and sent to the workflow selector logic. This process may repeat either indefinitely, until all objects and embedded objects are analyzed, or when a certain number of iterations/levels are reached. It is contemplated that rules may be set up for this process that can maximize threat analysis while avoiding computing overload by a large number of nested/embedded objects.

The analyzers generate results data that is sent to a correlation engine machine to generate score data (block 540). As shown above in more detail in the discussion of FIG. 2, the correlation engine logic can create score data based on a variety of factors including, but not limited to, analyzing the differences in the operation of the various analyzers. In some embodiments, the score data can be represented as a series of results or outputs of the various analyzers utilized in the analyzer logic. A pre-determined or dynamic threshold may be utilized that facilitates the determination of whether triage data generation should occur (block 545).

In certain embodiments where it is determined that a triage ticket should be generated the triage data can be then generated and passed into the alert logic which creates a triage ticket and reports it to the proper channels such that the newly generated triage ticket can be found and utilized by a human analyst for the purpose of determining the issues involved with the object threat detection in a more efficient manner. In certain embodiments, the triage data can be generated and utilized by the triage ticket generation logic to generate and distribute the triage ticket. In other embodiments, the triage data can be generated in response to every re-evaluation and utilized for other purposes not specific to triage ticket generation such as, but not limited to, analyzer testing, third party interactions, or other threshold evaluations.

In the foregoing description, the invention is described with reference to specific exemplary embodiments thereof. It will, however, be evident that various modifications and changes may be made thereto without departing from the broader spirit and scope of the invention as set forth in the appended claims.

Claims

1. An electronic device for automated triaging of cybersecurity threats, the electronic device comprising: a processor; anda memory communicatively coupled to the processor, the memory includes an object parsing logic configured to receive a submission message and parse the submission message into object data and context data;a workflow selector logic configured to receive the object data and process the object data to select at least one analyzer,an analyzer logic configured to (i) conduct analytics of object data using the at least one selected analyzer, and (ii) generate results data from the analytics,a correlation logic configured to receive the results data and generate score data based on the received results data, anda triage ticket generation logic configured to receive the score data and generate a triage ticket based on the score data surpassing one or more pre-defined thresholds.

2. A non-transitory storage medium including software that, when executed by a processor, conducts a supplemental analysis and classification of objects suspected of being falsely classified, the software comprising: an object parsing logic configured to receive a submission message and parse the submission message into object data and context data;workflow selector logic configured to receive at least the object data and process the object data to select one or more analyzers, wherein the one or more analyzers are configured to (i) conduct an analysis of the object data for maliciousness and (ii) generate results data based on the analysis of the object data;correlation logic configured to receive the results data and generate score data based on the received results data; andtriage ticket generation logic configured to receive the score data and generate triage data based on the score data.

3. The non-transitory storage medium of claim 2, wherein the workflow selector logic is further configured to receive the context data and process the context data to select the one or more analyzers.

4. The non-transitory storage medium of claim 3, wherein the context data comprises meta information about a cybersecurity threat detection system utilized by a source of the submission message and a classification associated with the object data.

5. The non-transitory storage medium of claim 2, wherein the workflow selector logic is configured to select the one or more analyzers in response to both the object data and the context data.

6. The non-transitory storage medium of claim 2, wherein the one or more analyzers include one or more behavioral analyzers each including a virtual machine to process the object data and settings of the one or more analyzers are configured based on the context data.

7. The non-transitory storage medium of claim 2, wherein the one or more analyzers emulate at least one software program with a specified software version with a plurality of software settings.

8. The non-transitory storage medium of claim 2, wherein the triage ticket generation logic is configured to generate the triage data including a triage ticket that includes information directed to a specific area of concern associated with classification of an object by the one or more analyzers.

9. The non-transitory storage medium of claim 2, wherein the software further comprising alert logic configured to automatically undertaking a remedial action based on the triage data.

10. A method for automated triaging of cybersecurity threats, the method comprising: receiving a submission message;parsing the submission message into object data and context data;receiving at least object data and processing the object data to select one or more analyzers, wherein the one or more analyzers are configured to (i) conduct an analysis of the object data for maliciousness and (ii) generate results data based on the analysis of the object data;receiving the results data and generating score data based on the received results data;receiving the score data and generating triage data based on the score data; andautomatically undertaking a remedial action based on the triage data to report a false positive result or a false negative result based on the score data.

11. The method of claim 10, wherein the receiving and processing of the object data further includes receiving the context data and processing the context data to select the one or more analyzers.

12. The method of claim 11, wherein the context data comprises meta information about a cybersecurity threat detection system utilized by a source of the submission message and a classification associated with the object data.

13. The method of claim 10, wherein the selecting of the one or more analyzers is conducted based on both the object data and the context data.

14. The method of claim 10, wherein the one or more analyzers include one or more behavioral analyzers each including a virtual machine to process the object data and settings of the one or more analyzers are configured based on the context data.

15. The method of claim 10, wherein the one or more analyzers emulate at least one software program with a specified software version with a plurality of software settings.

16. The method of claim 10, wherein the generating of the triage data includes generating a triage ticket that includes information directed to a specific area of concern associated with classification of an object of the object data by the one or more analyzers.

17. An electronic device comprising: a processor; anda non-transitory storage medium including software that, when executed by the processor, comprising: an object parsing logic configured to receive a submission message and parse the submission message into object data and context data;workflow selector logic configured to receive at least object data and process the object data to select one or more analyzers, wherein the one or more analyzers are configured to (i) conduct an analysis of the object data for maliciousness and (ii) generate results data based on the analysis of the object data,correlation logic configured to receive the results data and generate score data based on the received results data, andtriage ticket generation logic configured to receive the score data and generate triage data based on the score data.