Ransomware file modification prevention technique

Abstract

A computerized system and method to detect ransomware cyber-attacks is described. The approach entails analyzing the features associated with a file access event by a process operating on a computing device, to ascertain whether the process is associated with a ransomware cyber-attack.

Description

FIELD

Embodiments of the disclosure related to the field of cybersecurity. More specifically, embodiments of the disclosure relate to a system for detecting and preventing ransomware attacks before files are affected.

BACKGROUND

Ransomware is a fast-growing threat and is often distributed through the Internet. Ransomware is a type of malicious software (malware) that locks a user out of their device, applications, and/or data. Ransomware may operate by encrypting or otherwise making inaccessible files on an endpoint computer and, typically, alerting a user of the endpoint to the fact that they will not be able to access their files until a ransom is collected. Often ransomware will operate as a background process without any user interaction. Ransomware may target specific filetypes and/or content of high value to elicit payment from the affected user. Some ransomware may irreversibly render files inaccessible, despite ransom payment, making data irretrievable.

For example, ransomware may be downloaded and/or installed on a computer by “drive-by” download (silently installed in the background without user interaction while downloading webpages). Once installed on the endpoint, the user may have no further interaction with the ransomware until the contents of the computer are encrypted as part of the cyber-attack. At least once important data on the endpoint is encrypted, the endpoint user may be alerted by the ransomware to the potential for loss unless a ransom is paid.

Current solutions for ransomware prevention rely on detecting ransomware manipulating “honey files” and preventing that modification. Unfortunately, in the process ransomware is typically able to encrypt at least few of the legitimate files on the system. Moreover, these solutions are highly resource intensive and may fail to detect ransomware designed to ignore honey files. An improved ransomware detection system is needed to reliably protect endpoints from ransomware cyber-security attacks.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of this disclosure 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 block diagram of an architecture comprising a ransomware protection system (RPS) according to an embodiment of the invention;

FIG. 2 is a flowchart illustrating an exemplary method for detecting, preventing, and reporting ransomware attacks by the RPS of FIG. 1; and

FIG. 3 is an exemplary embodiment of a logical representation of the ransomware protection system of FIG. 1.

DETAILED DESCRIPTION

A ransomware prevention technique, limiting the effectiveness of cyber-attacks which encrypt or otherwise render inaccessible files, is provided. The ransomware prevention technique is implemented by a prevention system which (i) identifies a file access attempt (e.g., modify, delete, and/or move) by a process on a protected endpoint, (ii) intercepts the attempted file access by the process, (iii) determines if the attempted file access is permissible, and (iv) if permissible, allows the process to access the file while, if the access is impermissible, prevents access unless otherwise approved by the endpoint user.

More specifically, the ransomware prevention technique embodied in a ransomware prevention system monitors a protected endpoint system for attempted file accesses to one or more files during the execution of a process. One or more requested file accesses by an object or a computer program processing an object (referred to as a “file access events”), during runtime triggers the operation of the ransomware prevention system. An attempted file access request may be directed to any of a variety of files and file locations. These may include (a) any or all the files residing on (or accessible from) the protected endpoint system, (b) a specific type of file located on (or accessible from) the protected endpoint system (e.g., non-system files, excel files, etc.), and/or (c) files designated by the user or administrator of the protected endpoint, e.g., via a user interface. A monitoring logic is provided by the ransomware protection system to detect requests for file access (i.e. file access events). The ransomware prevention system (RPS) may operate by adapting or otherwise provisioning an application (e.g., Microsoft Word, etc.), with monitoring logic, and/or adapting or otherwise provisioning the operating system of the protected endpoint system to generate an analyzable attempted file access event (also referred to as a “file access event”) when triggered by a process during execution. In some embodiments, a processor-executable computer program (i.e. agent) may monitor one or more processes executing on the protected system and generate analyzable events when requests for file access are attempted by the process. A file access event may be generated in response to, but not limited to, an attempted file access, file modification, file move, and/or file delete of a protected file. In some embodiments, an attempted file access may further include attempted file accesses to create an incremental backup and/or “snapshot”. When the monitoring logic generates an analyzable event, the analyzable event is provided to a ransomware protection application for analysis.

When the ransomware prevention application (RPA) of the RPS generates an attempted file access event (i.e. an “analyzable event”), the RPA may generate an indicator to determine if the analyzable event is associated with known malicious (or benign) event indicators (i.e. a “known event indicator”). A known event indicator may be associated with at least a “block” (i.e. a blacklist) or “allow” (i.e. a whitelist) determination, the associated determination provided to a protection logic. The known event indicator and the respective determinations may be stored in an indicator store of the RPA. If the analyzable event is not associated with a known event indicator, the RPA may provide the analyzable event to a heuristics engine for further analysis.

The heuristics engine analyzes received characteristics and behaviors (collectively, “features”) associated with the analyzable event for suspiciousness. Features associated with the analyzable event may include, for example, the characteristics of the attempted file access event (e.g., filename and/or path of the file, the ID of the process requesting file access, etc.) as well as behaviors associated with the attempted file access event (e.g. file access request generated in response to a Graphical User Interface, “GUI”, interaction, processes requesting access to a several types of files, etc.) and behaviors related to the genealogy of the process requesting file access. The heuristics engine may examine the features associated with the analyzable event. The heuristics engine may also analyze objects associated with the received file access event, if available. If the analyzable event is determined to be suspicious by the heuristics engine, the engine may issue an alert to the user of the protected system and/or the security administrator of the system to take affirmative action to permit the process to proceed (or not) or, in some embodiments, provide an object associated with the analyzable event to a malware analysis system for further processing to determine if the object is malicious and its processing by the endpoint not permissible. In some embodiments, individual analyzable events may be analyzed for suspiciousness based on the features associated with the event while in alternative embodiments a plurality of analyzable events may be combined and the aggregated features of the combined analyzable events classified as suspicious by the classification engine.

The classification engine may receive from the heuristics engine the features detected and generate a classification of the analyzable event as being associated with “ransomware”, “suspicious”, or “benign”. The classification engine may generate the classification in response to determining a correlation of the features with known features of ransomware. The correlation of the features may be represented as a score, which when exceeding a first threshold may cause the event to be classified as being associated with ransomware based on machine learning and experiential knowledge. In some embodiments if the score does not exceed a second threshold, the analyzable event may be determined to be benign. The first and/or second threshold may be static (e.g., “factory-set”) or modifiable (i.e. in response to settings, endpoint user, etc.). In some embodiments the first and/or second threshold may be modifiable in response to an endpoint user role. If the classification engine determines that the analyzable event is associated with ransomware, the protection logic may be instructed to prevent further operations associated with the analyzable event (i.e. continued execution of the process associated with the analyzable event) while the reporting engine is instructed to report to a security administrator the ransomware event. When an analyzable event is determined to be associated with ransomware, the indicators associated with that analyzable event are stored in an indicator store of the RPA as known indicators associated with malware and/or ransomware. If the classification engine cannot determine the analyzable event is associated with ransomware or is benign (i.e. the ransomware score does not exceed the first threshold or second threshold), the analyzable event may be classified as “suspicious”.

In response to determining the analyzable event is suspicious the RPA may, in some embodiments, associate the analyzable event with a permissible operation and allow the operation to proceed; otherwise if the operation is not permissible, the operation may be prevented and an alert for the user and security administrator are generated and issued. In some embodiments, the RPA may enable the user to permit the operation for a suspicious analyzable event not previously known to be permissible via a user interaction with a user interface. If a suspicious analyzable event is deemed by the user to be permissible, its operation will be allowed and indicators associated with the analyzable event will be stored in the known indicator store and associated with benign (if the analyzable event is impermissible the indicator for the analyzable event will be associated with malware and/or ransomware in the known indicator store) events. The known indicator store may be based on characteristics of the protected file (e.g. file type). In some embodiments, in response to a suspiciousness determination of the analyzable event, the RPA may alert the user to determine whether the process generating the analyzable event is allowed to continue. If an attempted file access event is not determined to be permissible, the process initiating the analyzable event cannot receive the requested access to the file and the process will not continue execution (execution will be ended by the protection logic of the RPS).

The ransomware prevention technique described herein prevents ransomware cyberattacks by preventing their file operations. By analyzing features associated with attempted file accesses the system is able to identify suspicious operations and prevent ransomware from effecting operation on the protected system.

Terminology

In the following description, certain terminology is used to describe features of the invention. For example, in certain situations, the term “logic” may be representative of hardware, firmware and/or software that is configured to perform one or more functions. As hardware, logic may include circuitry having data processing or storage functionality. Examples of such circuitry may include, but are not limited or restricted to a microprocessor, one or more processor cores, a programmable gate array, a microcontroller, a controller, an application specific integrated circuit, wireless receiver, transmitter and/or transceiver circuitry, semiconductor memory, or combinatorial logic.

The term “process” may include an instance of a computer program (e.g., a collection of instructions, also referred to herein as an application). In one embodiment, the process may be included of one or more threads executing concurrently (e.g., each thread may be executing the same or a different instruction concurrently).

The term “processing” may include execution of a binary or script, or launching an application in which an object is processed, wherein launching should be interpreted as placing the application in an open state and, in some implementations, performing simulations of actions typical of human interactions with the application. For example, the application, an internet browsing application, may be processed such that the application is opened and actions such as “visiting” a website, downloading website pages, scrolling the website page, and activating a link from the website are performed.

The term “object” generally refers to a collection of data, whether in transit (e.g., over a network) or at rest (e.g., stored), often having a logical structure or organization that enables it to be categorized or typed for purposes of analysis. During analysis, for example, the object may exhibit a set of expected and/or unexpected features and, during processing, a set of expected and/or unexpected behaviors, which may evidence the presence of malware and potentially allow the object to be categorized or typed as malware. For example, an unexpected behavior of an object may include the generation of additional objects by an object being processed. In one embodiment, an object may include a binary file that may be executed within a virtual machine. Herein, the terms “binary file” and “binary” will be used interchangeably.

The term “file” is used in a broad sense to refer to a set or collection of data, information or other content used with a computer program. A file may be accessed, opened, stored, manipulated or otherwise processed as a single entity, object or unit. A file may contain other files, and may contain related or unrelated contents or no contents at all. A file may also have a logical format, and/or be part of a file system having a logical structure or organization of plural files. Files may have a name, sometimes called simply the “filename,” and often appended properties or other metadata. There are many types of files, such as data files, text files, program files, and directory files. A file may be generated by a user of a computing device or generated by the computing device. Access and/or operations on a file may be mediated by one or more applications and/or the operating system of a computing device. A filesystem may organize the files of the computing device of a storage device. The filesystem may enable tracking of files and enable access of those files. A filesystem may also enable operations on a file. In some embodiments the operations on the file may include file creation, file modification, file opening, file reading, file writing, file closing, and file deletion.

The term “feature” may be understood to refer, collectively, to characteristics of an object that may be detected statically and behaviors manifested in response to the processing of the object. Characteristics may include information about the object captured without requiring execution or “running” of the object. For example, characteristics may include metadata associated with the object, including, anomalous formatting or structuring of the object. Features may also include behaviors, where behaviors include information about the object and its activities captured during its execution or processing. Behaviors may include, but are not limited to, attempted outbound communications over a network connection or with other processes (e.g. the operating system, etc.), patterns of activity or inactivity, and/or attempts to access system resources.

The term “network device” may be construed as any intelligent electronic device with the capability of connecting to a network. Such a network may be a public network such as the internet or a private network such as a wireless data telecommunication network, wide area network, a type of local area network (LAN), or a combination of networks. Examples of a network device may include, but are not limited or restricted to a network appliance, laptop, mobile phone, or server.

The term “ransomware,” as described above, is a type of malicious software that is used to encrypt or otherwise prevent access to important information on a computer until payment (or “ransom”) is paid to the ransomware owner. Ransomware may rely on a network connection to receive confirmation of ransom payment while some ransomware requires entry of an unlock code that may be provided by the Ransomware owner once the ransom is paid. Ransomware may be time bound, deleting content from the computer or rendering the content permanently inaccessible if ransom is not paid before expiration of the ransomware.

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.

Ransomware Protection System

FIG. 1 is an exemplary block diagram of an exemplary architecture of a ransomware protection system (RPS) 100 connected to an enterprise network 102. The RPS 100 includes at least an operating system 110, application(s) 120, and a ransomware protection application 130, the ransomware protection application further comprising at least an indicator scanner 140, an indicator store 145, a classification engine 160, and a protection logic 175. The RPS may be operationally located as a component of a real endpoint (e.g. a user operated computer, server, mobile device, etc.) or virtual endpoint (e.g., a virtual machine, etc.). In some embodiments the operating system 110 may be adapted to include an operating system monitoring logic 115 and/or the application(s) 120 may be adapted to include an application monitoring logic 125. In some embodiments the ransomware protection application 130 may further include a heuristics engine 150 and a user interaction logic 170.

Generally speaking, the ransomware prevention system (RPS) 100 may be implemented as one or more components executed by at least one electronic computing device where each includes physical hardware comprising hardware processor(s), network interface(s), a memory, and a system interconnect as shown in FIG. 3. Accordingly, each of the components of the RPS shown in FIG. 1 may be implemented as one or more computer programs executable by the one or more processors and stored in memory. The RPS may operate as processor executable software by coordinating the operation of an operating system monitoring logic, an application monitoring logic 125, and a ransomware prevention application (RPA) 130.

The RPA 130 may include an indicator scanner 140, an indicator store 145, a classification engine, a protection logic 175 and a reporting engine 180. In some embodiments, the RPA 130 may further include a heuristics engine 150 and/or a protection logic 175. The RPA receives from the monitoring logic, of the RPS 100, operating on a user-operated electronic computing device (i.e. an endpoint), one or more file access events (“analyzable events”). The RPA 130 receives analyzable events, the analyzable events including features associated with the attempted file access for analysis to determine if the analyzable event is associated with a ransomware cyber-attack. The RPA 130 provides the analyzable event, and associated features, the indicator scanner 140 for analysis.

The indicator scanner 140 receives features associated with attempted file accesses generates indicators associated with the features, and compares the features with known indicators (e.g., known indicators may comprise known characteristics and/or known behaviors of malware) to determine if the analyzable event associated with the indicators are known to be associated ransomware (or malicious behavior) or are known to be benign. The received analyzable events may be provided by the monitoring logic of the RPS 100. The monitoring logic may be adapted to operate with the operating system of the RPS 100, embodied as operating system monitoring logic 115, and/or in one or more applications 120, embodied as application monitoring logic 125. The known indicators are each associated with a previously encountered analyzable event known to be “benign” or “ransomware”. In some embodiments, the indicator scanner 140 may be configured with a whitelist (indicators determined to be benign) and a blacklist (indicators determined to be associated with ransomware cyber-attacks).

The heuristics engine 150 applies heuristics and/or probability analysis to determine if the received file access events might include the hallmarks of ransomware. The heuristics engine 150 applies heuristics to the features of received events such as patterns of access, processing, etc. to determine a probability of being associated with ransomware. In some embodiments, the heuristics engine 150 may analyze the object indicated as a source of the file access event in coordination with a local malicious object analysis engine (not shown) or in coordination with remotely located malicious object analysis engines. The malicious object analysis engines may be located within an enterprise proprietary network or via an internet connect cloud network. The RPS 100 may provide to the malicious object analysis engine an object associated with the suspicious analyzable event and receive information associated with the maliciousness of the object. In some embodiments, the RPS 100 may receive an indicator associated with the classification to be retained in the indicator store 145. In some embodiments, the RPA 130 may prevent operation of the analyzable event until a response is received from the malicious object analysis engine.

When the determination of maliciousness of the attempted file access request is to be made by a separate service (i.e. a malicious object analysis engine), the RPA may temporarily determine the process requesting the attempted file access is to be treated as malicious (with the determination associated with features of the attempted file access event not stored in the known indicator store) and activate the protection logic, ending further execution of the process by the endpoint. Upon receipt from the separate service of an positive determination, by the separate malicious object analysis engine, of maliciousness of the process related to the attempted file access event, the indicators associated with the event are stored in the known indicator store in association with the determination and execution of the process continues to be blocked.

The heuristics engine 150 may be adapted to perform comparisons of an event and/or feature patterns under analysis against one or more pre-stored (e.g., pre-configured and/or predetermined) attack patterns stored in memory. The heuristics engine 150 may also be adapted to identify deviations in user file access patterns, the deviations determined through observation of user practices and experiential learning. Identified deviations from patterns of user file access are often a feature of ransomware cyber-attacks. The heuristics engine 150 may, in response to identifying patterns associated with ransomware generate a score indicative, to some level of probability, often well less than 100%, that the pattern is indicative of ransomware. The heuristics engine 150 may include a scoring logic to correlate one or more features, patterns of features of analyzable events associated with potential ransomware with known ransomware to generate a ransomware score. The generated ransomware score may be provided to the classification engine 160 for classification of the file access event as being associated with “ransomware” or “benign”.

The classification engine 160 receives from the heuristics engine 150 information associated with a likelihood of the analyzable event being associated with ransomware. Where the classification engine 160 receives from the heuristics engine 150 a correlation of the analyzable event and features with known ransomware, the classification engine may determine a score by weighting each correlation with each feature separately, and combining the separate scores using weights based on the experiential knowledge of the system. The classification engine 160 may be configured to use the scoring information provided by the heuristics engine 150 to classify an analyzable event as being associated with ransomware or benign. In one embodiment, when the score is above a first threshold, the classification engine 160 may generate an alert that the analyzable event is associated with ransomware. When the score is greater than a second threshold but lower than the first threshold, the analyzable event may be provided to the user interaction logic 170 for a user (or security administrator) to determine if the analyzable event should be permitted to continue operation (the analyzable event should not be associated with malware and/or ransomware). When the score is less than the second threshold, the classification engine 160 may determine no further analysis is needed and the process generating the analyzable event may continue operation (i.e., the event is benign). The one or more thresholds may be fixed, modified by a security administrator, and/or modified based on operating conditions of the electronic device (for example, if an electronic device is experiencing repeated events or attempted access or executions by certain processes sharing a set of features, etc.). In some embodiments, the classification engine 160 may use only the correlation information provided by the heuristics engine 150.

The protection logic 175 receives a determination, from the classification engine 160 and/or the optional user interaction engine 170, of a ransomware cyber-attack in response to a file access event. The protection logic 175 may coordinate with corresponding protection logic adapted for the operating system 110 or in one or more application(s) 120. In some embodiments, the protection logic may coordinate the respective monitoring logic of the operating system 110 or of the one or more application(s) 120 to prevent the continued operation of the ransomware process. The protection logic 175 may direct the operating system 110 to end a process determined to be associated with a ransomware cyber-attack by invoking a system function. In some embodiments, the protection logic 175 may be configured to intercept system or application functions associated with ransomware cyber-attack and block their continued processing. The reporting engine 180 is adapted to receive information from the classification engine 160 and generate alerts that identify to a user, administrator, and/or a security analyst the likelihood of a ransomware cyber-attack associated with the file access event (also known as an analyzable event). In some embodiments, the alert may identify the object associated the file access event which has been associated with the ransomware cyber-attack. Other additional information regarding the ransomware object may optionally be included in the alerts. For example, an alert may include information associated with the affected user, information associated with the object, and targeted files (based features associated with the file access event).

The user interaction logic 170 provides for and receives user input by the user of the endpoint device. The user input received by the user interaction logic 170 may include instructions to permit, or not, the file access prompted by the process. The user interaction logic may present to the user of the endpoint device the capability to permit (or not) the file access on a “one-time” basis, on a continuous basis, and/or for specifically identified files or sets of files bearing certain characteristics (e.g. a shared filetype, creator, directory, etc.). The availability of certain capabilities to permit (or not) may be limited based on information received by the heuristics engine indicating a level of suspiciousness. In some embodiments, endpoint device user instructed permission grants may be provided to the reporting engine 180 and an alert issued to the system administrator. If an end endpoint device user does not permit file access permission to a process, the instruction is provided to the protection logic 175.

The RPA 130 operates by coordinating with monitoring logic of the endpoint device configured to monitoring the access of files on the endpoint (i.e. the protected device). File accesses on the endpoint may be related to a legitimate process (i.e. a user initiated process or an automatic and/or scheduled process effected during the normal operation of the endpoint) or an illegitimate process (i.e. operating to effect otherwise unsanctioned modifications to files on the system). The operating system monitoring logic 115, application monitoring logic 125, and ransomware protection application 130 may be included in the application during activation of the ransomware protection system 100. The application monitoring logic 125 may be configured to detect attempted file accesses by an application 120 and coordinate with the RPA 130 via communication of an event. The application monitoring logic 125 may detect attempted file accesses by monitoring the functions of the application (i.e. application programming interface (API) function calls, operating system function calls, etc.) during execution. If the application monitoring logic 125 detects an attempted access via a monitored function the monitoring logic generates an event, including features associated with the application 120 and its attempted file access, and issues the event to the RPA 130. Similarly, the operating system monitoring logic 115 monitors the processes associated with the execution of the operating system 110. The operating system monitoring logic 115 is configured to detect attempted file accesses by monitoring operating system function calls. The operating system monitoring logic 115, adapted to coordinate with the operating system 110, may collect information associated with the operating system function calls monitored by the monitoring logic (i.e. calling process, associated processes, owner of the process, function arguments, etc.).

The RPA coordinates with the monitoring logic of the electronic device (endpoint) to determine if illegitimate processes are operating on a file of the device. For example, in some embodiments, an operating system monitoring logic 115 is operates with the operating system 120 of the endpoint. By coordinating with the operating system 110 via the operating system monitoring logic 115, the RPA 130 may be alerted to a process attempting to access an operating system file (also referred to as a “system file”) of the endpoint. Similarly, if an application is running on the endpoint and attempts to access a file, application monitoring logic 125 embedded in the application may be activated.

The operating system monitoring logic 115 and the application monitoring logic 125 of the ransomware protection system 100 identify behaviors (i.e. “observed behaviors”) associated with the processing of an object which initiates a file access event (i.e. features associated with executing a process). The monitoring logics, in some embodiments, may also identify characteristics associated with the processing of the object initiating the file access event. The observed behaviors and identified characteristics are features associated with the file access event. During classification by the classification engine 160, features may be used to classify the even as part of a ransomware cyber-attack.

In some embodiments, the monitoring logic may alternatively or additionally, operate at a different level of the software stack—e.g. at a hypervisor level in a virtual runtime environment—to intercept or capture such requests for file access (which generate analyzable events). More specifically, an executing process may send an API call containing a file access request, which becomes an operating system call, which in some embodiments may become a hypercall. Detection of the file access event at the hypervisor level may enable the implementation of the RPA in a secure analysis environment for such embodiments.

The application monitoring logic 125 monitor the operation of the one or more applications 120. In some embodiments, the application monitoring logic 125 and/or the operating system monitoring logic 115 intercepts behaviors associated with the event for analysis by the ransomware protection application 130.

During processing of an object on an electronic device protected by the ransomware protection system 100, the operating system monitoring logic 115 application monitoring logic 125 identify behaviors associated with the ransomware cyber-attacks (e.g. attempted file accesses, etc.). Signaling from application(s) 120 may be monitored through intercept points (sometimes referred to as “hooks”) to certain software calls (e.g., Application Programming Interface “API” call, library, procedure, function, or system call). The operations of the application may be monitored an intercept point (herein sometimes referred to as “instrumentation”) in code closely operating with the application to detect a certain type of activity, which may include an activity prompting a particular software call. In some embodiments, the monitoring logic may be embodied as hooks or instrumentation associated with calls to an operating system 120 function. The observed behavior associated with the processing of the object may be provided to the ransomware protection application 130 for analysis and determination if the feature is associated with a ransomware cyber-attack.

The RPA 130 may operate on an enterprise network 102 connected electronic device, such as a user operated endpoint computer. The RPA may communicate through the enterprise network 102 via a communication interface 305. The communication interface 305 may connect to the enterprise network 102 via a direct network interface or indirectly via a virtual private network. The enterprise network 102 may connected to the endpoint ransomware protection 100 via a communication interface 305. The communication interface 305 may connect to the enterprise network 102 via a direct network interface or indirectly via a virtual private network.

Referring now to FIG. 2, a flowchart, illustrating an exemplary method for identifying ransomware operating on a computing device and preventing the cyber-attack by the ransomware protection system (RPS) 100. Each block illustrated in FIG. 2 represents an operation performed in the method 200 of detecting phishing cyber-attacks with the RPS 100. The method starts at step 205 and proceeds to step 210 wherein the RPS identifies a file access event via one or more of the monitoring logics (i.e. the operating system monitoring logic 115 and/or the application monitoring logic 125). The monitoring logics monitor the operation of application(s) 120 and/or the operating system 110 for processing of data which may be associated with a ransomware attack. The RPS 100 first identifies potential ransomware processes before analyzing to determine if the process is associated with ransomware. In some embodiments, a signal of a ransomware attack is a file access event. The determination of whether the event and associated features are part of a ransomware cyber-attack, is made by analyzing the event and associated features in a ransomware protection application 130 in step 220.

In step 230, the indicator scanner 140 receives the event and associated features for further analysis (i.e. the event and associated features having been allowed for further analysis by step 225. The indicator scanner 140 analyzes the received event and features and determines if they match an indicator associated with a known benign and/or ransomware process. In some embodiments the indicator scanner 140 may coordinate with an indicator store 145, the indicator store is communicatively coupled to the indicator scanner 140. In some embodiments, the indicator store 145 may be updated with updated indicators associated known benign and/or ransomware by coordinating communication via the communication interface 305 of the RPA 100. The communication interface 305 may be used to update the indicator store 145 via a connection to an enterprise network 102. In some embodiments the indicator store 145 may be modified in response to a determination of ransomware by the classification engine 160, the indicator store generating a new indicator in response to the classification and storing the generated indicator. If the indicator scanner 140 determines that the event and/or features detected by the monitoring logics is associated with a known indicator in step 235, the method proceeds to step 240. If the indicator scanner 140 does not determine the event and/or features are known (i.e. not associated with a known benign or ransomware sample) the system proceeds to step 250. In some embodiments the indicator scanner 140 may provide a determination of whether or not the event and/or features are associated with a known benign or ransomware event to the classification engine 160 of the RPA 130.

In step 240, the indicator scanner 140 relays to the classification engine if the known association (i.e. benign or ransomware) is permissible. If the indicator scanner determines that the event and/or features are not permissible the indicator scanner will indicate to the protection logic 175 the associated event should be prevented in step 245. Alternatively, if the indicator scanner determines the event and/or features are permissible, the event will be permitted proceed without further intervention in step 270. For example, if the indicator scanner identifies the features of the file access event detected by the monitoring logics to be associated with known ransomware the indicator scanner may determine, in step 240, that the file access event is not permissible and prevent its further processing in step 245. In some embodiments, the protection logic may receive from the classification engine 160 a determination as to whether or not the event and/or features are permissible or should be prevented by the protection logic 175 in response to determination of association with an indicator by the indicator scanner 140.

In step 245 the protection logic 175 of the RPA 130 prevents further processing of the ransomware. The protection logic 175 may coordinate with protection logic of the operating system 110 or in the application(s) 120 to effect prevention of ransomware on the RPS 100. The protection logic may indicate to the operating system 110 to stop processing objects associated with a ransomware process (as determined by the classification engine 160). In some embodiments the protection logic 175 may intercept and block processing of associated processes with a particular file access event. Similarly, the API functions of an application may be used to effect prevention by ending the requested file access. When the protection logic 175 is initiated, an alert is generated and issued to an administrator in step 275.

The process continues in step 275 wherein, the reporting engine 180 receives information from the protection logic 175 and generates alerts issued via the communication interface 305 that identify to an administrator the information associated with the detected ransomware cyber-attack. The information provided by the alert may include, in part, information associated with the originating process, as well as the user associated with the ransomware. Additional information regarding the ransomware object (the object associated with the identified ransomware cyber-attack) may optionally be included in the alerts. The reporting engine 180 may also provide connected network security systems with updated information regarding ransomware attacks and the associated file access event and/or features do not represent a ransomware cyber-attack, the reporting engine 180 may alert an administrator via an alert, while in alternative embodiments the reporting engine will not issue an alert. Once step 275 is complete, the ransomware prevention procedure concludes at step 280.

Meanwhile, if in step 240 the indicator scanner 140 determines that file access event is not permissible the protection logic is bypassed by the classification engine and the file access event is permitted to proceed and receive the required access without further intervention by the RPA 130. In some embodiments, the indicator scanner 140 provides the determination to the classification engine 160, while in other embodiments the classification engine receives a determination of whether or not the indicator is associated with ransomware or is benign and determines if the process is permissible.

If, in step 235, the indicator scanner 140 does not determine that the event and/or features associated with the event have been previously seen (i.e. are stored in the indicator store 145) the process proceeds to step 250 where the event and/or its associated features are provided to a heuristics engine 150 for further analysis. The heuristics engine 150 receives the event and/or features associated with the event and applies heuristics, rules or other conditions to the features to identify anomalous or unexpected behaviors which may be associated with ransomware. The analysis of the heuristics engine may yield an association between a set of one or more features and/or the event and a likelihood of association with ransomware. This likelihood of association with ransomware may be provided to a classification engine 160 for ransomware classification in step 260.

In step 260 the classification engine 160 receives from the heuristics engine 150 a likelihood of association with ransomware of the monitored features and/or file access event. In some embodiments the classification engine may receive an association with known ransomware or benign indicators from the indicator scanner 140. The classification engine 160 determines, based on the received information associated with each file access event whether or not it is associated with ransomware. If the classification engine 160 determines the event is associated with ransomware because the likelihood of association with ransomware exceeds a first threshold, the event will be determined and labelled as “ransomware” in step 265. Meanwhile, in some embodiments, if the likelihood of association with ransomware is below a second threshold, the event may be determined and labelled to be “benign”. If the association with ransomware is not labelled “benign” or “ransomware”, the event may, in some embodiments be labelled as “suspicious”. Suspicious events may be provided to a user interaction logic 170 for classification by a user. Based upon the user in response to an alert generated by the user interaction engine 170, the event will be classified as ransomware or benign. File access events determined to be benign by the classification engine 160 or the user interaction logic 170 may be permitted to receive access to the file requested in step 270. Meanwhile, in some embodiments, suspicious events may be prevented. Prevented events (including events classified as ransomware and in some embodiments, suspicious events) are provided to the reporting engine 180 in step 275 and in some embodiments to the protection logic 175. The process continues in step 275 as described above.

In step 220, the RPA receives from one or more monitoring logics the file access event and any associated features. The inventors recognize certain hallmarks of ransomware (or other malware rendering files maliciously encrypted, deleted or otherwise inaccessible) may be used as features for the classification of cyber-attacks as ransomware. The features may include, for example, details of the attempted file access (e.g., number of files accesses, file type of file(s) accessed, a process ID, process parent history, etc.), calling process history (i.e. is the calling process associated with another process, was the calling process directed by user-input, etc.), and/or whether there is an association with a user interface (i.e. did a user initiate the process via a Graphical User Interface which generated the analyzable event). Additional information which may be included as features are statistics related to the characteristics of the attempted file access event. The RPA 130 may use determine whether or not a file access event will be analyzed for ransomware protection based on the associated features of the object in step 220, alternatively, the procedure may proceed directly to step 230.

FIG. 3 is an exemplary embodiment of a logical representation of the ransomware protection system 100 of FIG. 1. The ransomware protection system 100, in an embodiment may include a housing, which is made entirely or partially of a hardened material (e.g., hardened plastic, metal, glass, composite or any combination thereof) that protects the circuitry within the housing, namely one or more processors 310 that are coupled to a communication interface 305 via a first transmission medium 307. The communication interface 305, in combination with a communication interface logic 320, enables communications with external network devices and/or other appliances via a network to receive updates for the ransomware prevention system 100. According to one embodiment of the disclosure, the communication interface 305 may be implemented as a physical interface including one or more ports for wired connectors. Additionally, or in the alternative, the communication interface 305 may be implemented with one or more radio units for supporting wireless communications with other electronic devices. The communication interface logic 320 may include logic for performing operations of receiving and transmitting one or more objects via the communication interface 305 to enable communication between the ransomware protection system 100 and network devices via a network (e.g., the internet) and/or cloud computing services.

The processor(s) 310 is further coupled to a persistent storage 315 via a second transmission medium 313. According to one embodiment of the disclosure, the persistent storage 315 may include a communication interface logic 320, an operating system 110, an operating system monitoring logic 115, application(s) 120, application monitoring logic 125, a ransomware protection application 130, an indicator scanner 140, a classification engine 160, a protection logic 175 and a reporting engine 180, as well as an optional heuristics engine 150 and/or a user interaction engine 170. When implemented as hardware, one or more of these logic units could be implemented separately from each other.

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. A system to detect and prevent ransomware cyber-attacks on a computer, the system comprising: one or more processors; anda storage device communicatively coupled to the one or more processors, the storage device to store a plurality of logic modules for execution by the one or more processors, the plurality of logic modules include an indicator scanner, a classification engine, a protection logic, an user interaction logic, and a reporting engine, wherein, the indicator scanner is configured to receive an analyzable event in response to at least one attempted file access by a process and extract features from the analyzable event relating to the at least one attempted file access and the process,the classification engine is configured to (i) receive the extracted features, (ii) compare the features with a set of known indicators associated with determinations of a cyber-attack to generate a maliciousness score based on a correlation between the extracted features and the set of known indicators, and (ii) classify the analyzable event as either (a) malicious when the maliciousness score exceeds a first threshold or (b) non-malicious when the maliciousness score fails to exceed a second threshold,the user interaction logic is configured to (i) receive the analyzable event in response to the maliciousness score having a value between the first threshold and the second threshold and (ii) provide a classification of the analyzable event to the classification engine where the classification is based on a determination by a user or administrator as to whether the analyzable event is malicious,the protection logic is configured to prevent continued execution of the process in response to the classification engine classifying the analyzable event as malicious, andthe reporting engine is configured to generate and issue an alert in response to the classification of the analyzable event as malicious.

2. The system of claim 1, wherein the plurality of logic modules further comprising monitoring logic that, when executed by the one or more processors, monitors one or more processes including the process, generates the analyzable event in response to the at least one attempted file access by the process, and provides the analyzable event to the indicator scanner analysis.

3. The system of claim 2, wherein the monitoring logic comprises at least one of an operating system monitoring logic interoperable with an operating system of the computer for monitoring for attempted access requests therefrom, and an application monitoring logic interoperable with at least one application operating on the computer for monitoring for attempted access requests therefrom.

4. The system of claim 1, wherein the analyzable event is deemed to be malicious when the analyzable event is associated with a cyber-attack.

5. The system of claim 1, wherein the protection logic is further configured to: extract from the analyzable event a process identifier (ID) as a first feature, and prevent execution of the process associated with the malicious analyzable event by signaling the processor to stop execution of the process identified by the process ID.

6. The system of claim 1, wherein the analyzable event is deemed to be malicious when the process is associated with a cyber-attack.

7. The system of claim 6, wherein the cyber-attack is a ransomware cyber-attack.

8. The system of claim 1, wherein the first threshold is greater in score value than the second threshold.

9. The system of claim 1, wherein the protection logic is operationally located in a hypervisor hosting at least one virtual environment provisioned as the computer and utilizing at least the protection logic and the indicator scanner.

10. The system of claim 1, wherein the reporting engine, when executed by the one or more processors, generates and issues the alert in response to receiving, from the protection logic, information specifying the process associated with the classification of malicious was prevented from continued execution.

11. The system of claim 1, wherein the reporting engine, when executed by the one or more processors, generates the alert via a graphical user interface for an administrator to classify the analyzable event as being associated with a malicious process in response to the comparison.

12. The system of claim 1, wherein the classification engine to compare the features with the set of known indicators being associated with known indicators of ransomware to classify the features of the analyzable event as ransomware or not.

13. The system of claim 1, wherein the one or more processors are one or more hardware processors.

14. A computerized method for detecting a ransomware cyber-attack on a computer, the method comprising: capturing one or more attempted file access requests by monitoring execution of a process by the computer and generating an analyzable event;extracting features associated with the analyzable event;examining the extracted features with features associated with known malicious indicators associated with ransomware or known benign indicators for classifying the analyzable event as a malicious event, the examining of the extracted features includes generating a maliciousness score based on a correlation of the extracted features with the known malicious indicators;classifying the analyzable event as either the malicious event when the maliciousness score satisfies a first threshold or a non-malicious event when the maliciousness score fails to satisfy a second threshold;classifying, by a user or administrator, whether the analyzable event corresponds to the malicious event or the non-malicious event when the maliciousness score has a value between the first threshold and the second threshold;preventing further operations by the process of the computer in response to the analyzable event being classified as a malicious event; andgenerating and issuing an alert to an administrator in response to classifying the analyzable event as a malicious event.

15. The method of claim 14, further comprising: requiring user input of a computer user as to maliciousness, in response to detecting a filetype associated with the event.

16. The system of claim 14, wherein the preventing of further operations by the process of the computer comprises extracting from the analyzable event a process identifier (ID) as a first feature, and preventing execution of the process associated with the malicious analyzable event by signaling the processor to stop execution of the process identified by the process ID.

17. The method of claim 14, wherein the examining of the exacted features associated with the known malicious indicators or features associated within known benign indicators comprises generating the maliciousness score in response to a correlation of the extracted with the known malicious indicators associated with ransomware, the maliciousness score being used to classify the analyzable event as ransomware.

18. The method of claim 17, wherein the analyzable event is determined to be ransomware in response to the maliciousness score exceeding a first threshold.

19. The method of claim 14, wherein the preventing further operations by the process of the computer is performed by protection logic operationally located in a hypervisor hosting at least one virtual environment provisioned as the computer.

20. The method of claim 14, wherein the generating and issuing of the alert to the administrator occurs in response to receiving information specifying the process being classified as the malicious event was prevented from continued execution.

21. A non-transitory computer readable medium including software that, upon execution by one or more processors, performs operations to determine a presence of ransomware, the medium comprising: an indicator scanner configured to receive an analyzable event in response to at least one attempted file access by a process and extract features from the analyzable event relating to the at least one attempted file access and the process;a classification engine configured to (i) receive the extracted features, (ii) compare the features with a set of known indicators associated with determinations of a cyber-attack to generate a maliciousness score based on a correlation between the extracted features and the set of known indicators, and (ii) classify the analyzable event as either (a) malicious when the maliciousness score exceeds a first threshold or (b) non-malicious when the maliciousness score fails to exceed a second thresholda user interaction logic configured to provide, based on an input by a user or administrator, a classification of the analyzable event to the classification engine, the classification being based on a determination by the user or administrator as to whether the analyzable event is malicious;a protection logic configured to prevent, in response to the classification engine or the user interaction logic classifying the analyzable event as malicious, continued execution of the process by the one or more processors; anda reporting engine configured to generate and issue an alert in response to the classification of the analyzable event as malicious.

22. The non-transitory computer readable medium of claim 21, wherein the protection logic is further configured to extract from the analyzable event a process identifier (ID) as a first feature, and prevent execution of the process associated with the malicious analyzable event by signaling the processor to stop execution of the process identified by the process ID.

23. The non-transitory computer readable medium of claim 21, wherein the analyzable event is deemed to be malicious when the process is associated with a cyber-attack.

24. The non-transitory computer readable medium of claim 23 wherein the cyber-attack is a ransomware cyber-attack.

25. The non-transitory computer readable medium of claim 21 wherein the protection logic is operationally located in a hypervisor hostig at least one virtual environment including the protection logic and the indicator scanner.

26. The non-transitory computer readable medium of claim 21, wherein the reporting engine, when executed by the one or more processors, generates and issues the alert in response to receiving, from the protection logic, information specifying the process associated with the classification of malicious was prevented from continued execution.