A system and method for determining one or more security policies

The system optimizes security policies by analyzing organizational assets and tools to balance security and productivity, addressing fragmented tool landscapes and improving overall cybersecurity posture.

WO2026234659A1PCT designated stage Publication Date: 2026-11-12RECLAIM SECURITY LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/IL2026/050183
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-08
Filing Date
2026-02-26
Publication Date
2026-11-12

AI Technical Summary

Technical Problem

Organizations face challenges in establishing coherent security policies across fragmented security tool landscapes, leading to overlapping or contradictory requirements, decreased productivity, and increased cybersecurity risks due to traditional approaches that prioritize stringent controls over business continuity.

Method used

A system and method for determining security policies that analyze organizational assets and security tools to identify asset groups with statistical similarity, predict productivity impacts, and balance security needs, using machine learning and heuristic techniques to optimize policy configurations.

Benefits of technology

The system enables organizations to create context-aware security policies that enhance protection while minimizing productivity disruptions by integrating security tools and adapting to organizational patterns and business needs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure IL2026050183_12112026_PF_FP_ABST
    Figure IL2026050183_12112026_PF_FP_ABST
Patent Text Reader

Abstract

A system for determination of one or more security policies, the system comprising a processing circuitry configured to perform a process including: obtain: (A) data associated with one or more assets of an organization, and (B) information of one or more controls, each of the controls is associated with a security tool of one or more security tools of the organization, and is: (i) associated with a capability of a respective security tool of the security tools to prevent one or more given assets of the assets from taking one or more actions, and (ii) associated with one or more security policies of the security policies, wherein each security policy is a configuration of the given security tool that effects a group of assets of the assets; perform the following, for at least one given control of the controls, and for at least one given action of the actions associated with the given control: find, based on the data, one or more given assets that are associated with the given action, giving rise to an associated asset group; cluster the assets into one or more asset groups, wherein at least one asset group of the asset groups comprises one or more assets of the assets that have statistical similarity, and wherein the statistical similarity is determined by analyzing the associated asset group; and determine, for at least one given asset group of the asset groups, a respective security policy, based on a prediction of a productivity impact of preventing the assets of the given asset group from preforming the given action; and perform an action based on the determined security policies.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] A SYSTEM AND METHOD FOR DETERMINING ONE OR MORE SECURITY POLICIES TECHNICAL FIELD

[0002] The present invention relates to the field of a system and method for determining one or more security policies.

[0003] BACKGROUND

[0004] The current organizational security policy landscape is characterized by increasing complexity due to the proliferation of numerous security tools designed to address specific threats. Organizations today deploy firewalls, antivirus software, Intrusion Detection Systems (IDSs), identity and access management solutions, and cloud security platforms, among many others. Each of these tools often requires its own distinct policies, configurations, and monitoring practices. As a result, organizations struggle to establish coherent, effective security policies across this fragmented tool landscape without negatively impacting employee productivity. These traditional security approaches force organizations to choose between strong security and business continuity.

[0005] For instance, a typical organization may use separate security tools for email filtering, endpoint protection, Data Loss Prevention (DLP), and network monitoring. Each security tool has its own set of controls and generates its own set of alerts, requiring specialized attention from security teams. This fragmentation makes security policy determination particularly challenging, as security policies optimized for one tool may inadvertently interfere with the functionality or usability of another. Consequently, various assets of the organization (e.g., networks, servers, computers, devices, employees and more) might find themselves subjected to overlapping or contradictory security requirements, leading to frustration, decreased efficiency and diminished productivity.

[0006] Traditional security approaches often exacerbate this issue by emphasizing stringent controls and restrictions aimed at minimizing risk at all costs. For example, many organizations traditionally impose restrictive web filtering policies that block social media platforms or cloud-based collaboration tools to prevent potential breaches or data leaks. However, these overly cautious measures can severely hinder businessproductivity, especially in environments where collaboration, remote work, and cloud services are integral to operational effectiveness.

[0007] This forced choice between robust security and smooth business operations creates a significant strategic dilemma. Organizations frequently find themselves having to either weaken their security posture to enhance business continuity or accept considerable productivity losses to maintain stringent security standards. Many times, these decisions are taken without understanding the full implication on the productivity of the organization and on the amount of risk taken. There is a lack of transparency to what are the tradeoffs for such decisions. For example, overly aggressive DLP rules may prevent employees from sharing files via essential collaboration platforms like Microsoft Teams or Google Workspace, driving teams to seek out unauthorized, less secure methods of communication, thereby ironically increasing the organization's overall risk profile.

[0008] A non-limiting example can be of a manufacturing company that needs to implement application control policies to prevent malware without disrupting critical engineering tools. Traditional endpoint protection solutions often utilize rigid allowlist and blocklist / denylist methods, potentially blocking legitimate, specialized applications required by engineering teams, such as Computer Aided Design (CAD) software or proprietary firmware update utilities. Overly aggressive application controls may mistakenly categorize these tools as unauthorized or risky, disrupting manufacturing workflows and significantly impacting productivity. Consequently, engineering teams might bypass security controls or adopt unauthorized workarounds, inadvertently increasing cybersecurity risks.

[0009] An additional non-limiting example can be of a healthcare organization that needs to implement stronger authentication requirements without disrupting clinical workflows, especially during emergency situations. Traditional security policies, such as enforcing Multi-Factor Authentication (MFA) for every login, might provide robust protection but could critically delay clinical staff when accessing Electronic Health Records (EHR) in time-sensitive scenarios, such as: emergency room admissions or urgent patient care interventions. These stringent authentication measures, while essential for protecting patient data and complying with regulatory standards like Health Insurance Portability and Accountability Act (HIPAA), can inadvertently introduce risks to patient safety by obstructing rapid access to critical medical information.Another non-limiting example is an implementation of a Cloud Access Security Broker (CASB) at a financial services firm. This organization faces the challenging task of securing its cloud applications to comply with stringent regulatory requirements and protect sensitive client data. However, overly restrictive CASB policies could unintentionally block legitimate activities, such as real-time financial transactions or collaborative client meetings facilitated via cloud-based platforms. The challenge is magnified by the need to finely tune policies that enforce robust data protection controls without interrupting legitimate business operations, highlighting the difficulty in striking an optimal balance between strong security measures and ongoing business productivity.

[0010] An additional non-limiting example illustrates the challenge of determining effective security policies when considering the complete attack-chain across multiple organizational assets. Modern cyber-attacks rarely target a single security layer, but instead execute through multi-step flows that traverse various organizational systems and security tools. In this non-limiting example, a sophisticated attack might begin with a phishing email containing a malicious Portable Document Format (PDF) file, which then executes on a Windows endpoint device, establishes command-and-control network connections, and ultimately exfiltrates sensitive data. Traditional security approaches often lead organizations to implement redundant, maximum-strength controls at every layer without understanding how these controls interrelate across the attack chain. This can result in unnecessary productivity disruptions. For example, if an organization has implemented rigorous email security that effectively blocks all incoming malicious PDF files at the email gateway, the need for equally stringent PDF blocking policies on endpoint devices becomes less critical for addressing this specific attack vector. The risk from PDFs arriving via external email has already been mitigated at the email infrastructure level. However, security teams frequently lack visibility into these interdependencies between security layers, leading to overzealous policy implementation that creates productivity bottlenecks without proportional security benefits. The challenge is further complicated when different teams manage different security tools, often making decisions in isolation without a comprehensive understanding of the entire security ecosystem. This siloed approach prevents organizations from developing a cohesive security strategy that optimizes both protection and productivity across the full attack chain.Without tools that provide insight into how security controls interact across multiple steps of an attack flow, organizations struggle to make informed decisions about where strong controls are necessary versus where they can be more measured to preserve productivity. This comprehensive, attack-chain perspective is crucial for determining security policies that provide effective protection while minimizing unnecessary business disruption.

[0011] These challenges in determining organizational security policies are substantial. Organizations must continuously balance multiple, overlapping tools and increasingly sophisticated threats, all while managing employee productivity and satisfaction. This ongoing challenge underscores the importance of developing security policies determination tools that instruct the organization on the security risks and productivity impact associated with these policy changes. These tools should be both integrated and flexible. Ultimately, organizations that successfully harmonize robust security practices with seamless business continuity are likely to achieve a competitive advantage, illustrating the pressing need to rethink traditional approaches in favor of adaptive, business and context-aware security policies.

[0012] Thus, there is a need for a novel technique for a system and method for determining business-aware security policies that maximizes protection while preserving productivity.

[0013] GENERAL DESCRIPTION

[0014] In accordance with a first aspect of the presently disclosed subject matter, there is provided a system for determination of one or more security policies, the system comprising a processing circuitry configured to perform a process including: obtain: (A) data associated with one or more assets of an organization, and (B) information of one or more controls, each of the controls is associated with a security tool of one or more security tools of the organization, and is: (i) associated with a capability of a respective security tool of the security tools to prevent one or more given assets of the assets from taking one or more actions, and (ii) associated with one or more security policies of the security policies, wherein each security policy is a configuration of the given security tool that effects a group of assets of the assets; perform the following, for at least one given control of the controls, and for at least one given action of the actions associated with the given control: find, based on the data, one or more given assets that are associated withthe given action, giving rise to an associated asset group; cluster the assets into one or more asset groups, wherein at least one asset group of the asset groups comprises one or more assets of the assets that have statistical similarity, and wherein the statistical similarity is determined by analyzing the associated asset group; and determine, for at least one given asset group of the asset groups, a respective security policy, based on a prediction of a productivity impact of preventing the assets of the given asset group from preforming the given action; and perform an action based on the determined security policies.

[0015] In some cases, the processing circuitry is further configured to: apply a given security policy of the security policies on at least one given control of the controls, based on an Information Technology (IT) feasibility of the security tool associated with the given control.

[0016] In some cases, the apply includes merging at least two of the determined security policies associated with the respective given control into a merged determined security policy and wherein the action is performed on the merged determined security policies.

[0017] In some cases, the merging of the at least two of the determined security policies into a merged security policy is based on a security value of the action associated with the respective security policy.

[0018] In some cases, the security value of a given action is determined based on one or more Tactics, Techniques and Procedures (TTPs) that involve the given action.

[0019] In some cases, the TTPs are based on MITRE.

[0020] In some cases, the TTPs are weighted in accordance with one or more characteristics of the organization.

[0021] In some cases, the TTPs are weighted in accordance with a risk level.

[0022] In some cases, the merging of the at least two of the determined security policies into a merged security policy is performed utilizing a heuristic technique.

[0023] In some cases, the apply includes selecting at least one of the determined security policies associated with the respective given control into a selected determined security policy and wherein the action is performed on the selected determined security policies.

[0024] In some cases, prior to the determination of the statistical similarity, the data associated with the assets of the performing asset group is pre-processed.

[0025] In some cases, the pre-processing is performed by removing one or more events associated with the given action.In some cases, the determination of the statistical similarity of the assets is preformed utilizing a machine learning model.

[0026] In some cases, the machine learning model is capable of receiving: the pre-processed data associated with a given asset of the assets, and determining an embedding vector representative of the asset, wherein the machine learning model is trained utilizing the pre-processed data associated with the assets of the performing asset group.

[0027] In some cases, the clustering of the assets into one or more asset groups is preformed utilizing the embedding vectors representative of the respective assets.

[0028] In some cases, the determination of the respective security policy is preformed utilizing one or more rules balancing a calculated productivity impact on the assets of the given asset group and security needs of the organization.

[0029] In some cases, the rules reflect a given level of allowed false positives.

[0030] In some cases, the allowed false positives are distributed between one or more of: (a) the assets in accordance with organizational information of the respective assets, (b) the controls in accordance with organizational information of the respective controls, or (c) the actions in accordance with organizational information of the respective actions.

[0031] In some cases, the finding of the one or more given assets that are associated with the given action is performed within a time window.

[0032] In some cases, the one or more assets can include one or more of: a server, a computer, a laptop, a security tool, a network component, a smart device, a user, an application, a container, a virtual machine, a cloud service, an loT device, a mobile device, a database, an API endpoint, a software service, a data repository, data, a file, a document, a non-human identity, a service account, a machine identity, a digital certificate, a key, a token, a domain, a subnet, or a network segment.

[0033] In some cases, the obtained data can include one or more of: logs of events associated with the respective asset, organizational information of the respective asset, information from external sources associated with the respective asset, proxy data derived from one or more additional assets related to the respective asset wherein direct visibility is limited, indirect logs from one or more interconnected systems, inferences drawn from absence of expected data, behavioral patterns identified through partial process observations, correlative data from sequential event analysis wherein observed initial steps without subsequent steps indicate incomplete execution chains, or synthesized visibility constructed from fragmented or incomplete monitoring data.In some cases, the processing circuitry is further configured to: simulate, before performing the action, a deployment of at least one of the determined security policies to at least one security tool of the security tools; and predict an operational impact of the simulated deployment on one or more assets of the assets.

[0034] In some cases, the processing circuitry is further configured to: simulate, before performing the action, a deployment of at least one of the determined security policies to at least one security tool of the security tools; and quantify an expected risk reduction of the simulated deployment on one or more assets of the assets.

[0035] In some cases, the action is one or more of: a recommendation to a user of the system, an automatic deployment of at least one of the determined security policies to at least one security tool of the security tools, a semi-automatic deployment of at least one of the determined security policies to at least one security tool of the security tools requiring human intervention, or a simulated deployment of at least one of the determined security policies to at least one security tool of the security tools.

[0036] In some cases, the associated asset group comprises of assets that have been found to: (a) perform the given action, or (b) not perform the given action.

[0037] In some cases, the finding, based on the data, of one or more given assets that are associated with the given action can be directly from the data or indirectly by analyzing proxy data associated with other assets, other than the given assets.

[0038] In some cases, the obtained data is enriched from additional sources.

[0039] In some cases, the processing circuitry is further configured to: asses the results of performing the action on the organization; and determine, based on the assessment, one or more changes to the security policies.

[0040] In some cases, the prediction of productivity impact comprises quantitative assessment of one or more of:

[0041] (a) workflow interruptions measured by increases in task completion time;

[0042] (b) user experience degradation measured by: (i) additional authentication steps, (ii) access request frequency, or (iii) interface interactions, process delays resulting from security policy enforcement;

[0043] (c) resource utilization changes, including: CPU, memory, network, or storage consumption;

[0044] (d) application performance metrics, including: response time, throughput, or availability;(e) collaboration friction introduced between one or more entities of the organization; (f) increased IT support requests;

[0045] (g) increased security exception applications; or

[0046] (h) increased opportunity costs arising from restricted access to required resources.

[0047] In some cases, the processing circuitry preforms at least part of the process repeatedly.

[0048] In some cases, the obtain includes obtaining new data associated with one or more of the assets of the organization and wherein the processing circuitry is further configured to: determine, based on the new data, new security policies for one or more of the assets; detect a security policy drift of one or more of the assets, being assets that have a difference above a threshold between the respective determined security policy and the respective new security policy; and automatically address the security policy drift by amending the respective determined security policy utilizing the new respective determined security policy.

[0049] In accordance with a second aspect of the presently disclosed subject matter, there is provided a method for determination of one or more security policies, the method comprising: performing a process including: obtaining, by a processing circuitry: (A) data associated with one or more assets of an organization, and (B) information of one or more controls, each of the controls is associated with a security tool of one or more security tools of the organization, and is: (i) associated with a capability of a respective security tool of the security tools to prevent one or more given assets of the assets from taking one or more actions, and (ii) associated with one or more security policies of the security policies, wherein each security policy is a configuration of the given security tool that effects a group of assets of the assets; performing, by the processing circuitry, the following, for at least one given control of the controls, and for at least one given action of the actions associated with the given control: finding, by the processing circuitry, based on the data, one or more given assets that are associated with the given action, giving rise to an associated asset group; clustering, by the processing circuitry, the assets into one or more asset groups, wherein at least one asset group of the asset groups comprises one or more assets of the assets that have statistical similarity, and wherein the statistical similarity is determined by analyzing the associated asset group; and determining, by the processing circuitry, for at least one given asset group of the asset groups, a respective security policy, based on a prediction of a productivity impact of preventing the assets ofthe given asset group from preforming the given action; and performing, by the processing circuitry, an action based on the determined security policies.

[0050] In some cases, the method further comprising: applying, by the processing circuitry, a given security policy of the security policies on at least one given control of the controls, based on an Information Technology (IT) feasibility of the security tool associated with the given control.

[0051] In some cases, the apply includes merging at least two of the determined security policies associated with the respective given control into a merged determined security policy and wherein the action is performed on the merged determined security policies.

[0052] In some cases, the merging of the at least two of the determined security policies into a merged security policy is based on a security value of the action associated with the respective security policy.

[0053] In some cases, the security value of a given action is determined based on one or more Tactics, Techniques and Procedures (TTPs) that involve the given action.

[0054] In some cases, the TTPs are based on MITRE.

[0055] In some cases, the TTPs are weighted in accordance with one or more characteristics of the organization.

[0056] In some cases, the TTPs are weighted in accordance with a risk level.

[0057] In some cases, the merging of the at least two of the determined security policies into a merged security policy is performed utilizing a heuristic technique.

[0058] In some cases, the apply includes selecting at least one of the determined security policies associated with the respective given control into a selected determined security policy and wherein the action is performed on the selected determined security policies.

[0059] In some cases, prior to the determination of the statistical similarity, the data associated with the assets of the performing asset group is pre-processed.

[0060] In some cases, the pre-processing is performed by removing one or more events associated with the given action.

[0061] In some cases, the determination of the statistical similarity of the assets is preformed utilizing a machine learning model.

[0062] In some cases, the machine learning model is capable of receiving: the pre-processed data associated with a given asset of the assets, and determining an embedding vector representative of the asset, wherein the machine learning model is trained utilizing the pre-processed data associated with the assets of the performing asset group.In some cases, the clustering of the assets into one or more asset groups is preformed utilizing the embedding vectors representative of the respective assets.

[0063] In some cases, the determination of the respective security policy is preformed utilizing one or more rules balancing a calculated productivity impact on the assets of the given asset group and security needs of the organization.

[0064] In some cases, the rules reflect a given level of allowed false positives.

[0065] In some cases, the allowed false positives are distributed between one or more of: (a) the assets in accordance with organizational information of the respective assets, (b) the controls in accordance with organizational information of the respective controls, or (c) the actions in accordance with organizational information of the respective actions.

[0066] In some cases, the finding of the one or more given assets that are associated with the given action is performed within a time window.

[0067] In some cases, the one or more assets can include one or more of: a server, a computer, a laptop, a security tool, a network component, a smart device, a user, an application, a container, a virtual machine, a cloud service, an loT device, a mobile device, a database, an API endpoint, a software service, a data repository, data, a file, a document, a non-human identity, a service account, a machine identity, a digital certificate, a key, a token, a domain, a subnet, or a network segment.

[0068] In some cases, the obtained data can include one or more of: logs of events associated with the respective asset, organizational information of the respective asset, information from external sources associated with the respective asset, proxy data derived from one or more additional assets related to the respective asset wherein direct visibility is limited, indirect logs from one or more interconnected systems, inferences drawn from absence of expected data, behavioral patterns identified through partial process observations, correlative data from sequential event analysis wherein observed initial steps without subsequent steps indicate incomplete execution chains, or synthesized visibility constructed from fragmented or incomplete monitoring data.

[0069] In some cases, the method further comprising: simulating, by the processing circuitry, before performing the action, a deployment of at least one of the determined security policies to at least one security tool of the security tools; and predicting, by the processing circuitry, an operational impact of the simulated deployment on one or more assets of the assets.In some cases, the method further comprising: simulating, by the processing circuitry, before performing the action, a deployment of at least one of the determined security policies to at least one security tool of the security tools; and quantifying, by the processing circuitry, an expected risk reduction of the simulated deployment on one or more assets of the assets.

[0070] In some cases, the action is one or more of: a recommendation to a user of the system, an automatic deployment of at least one of the determined security policies to at least one security tool of the security tools, a semi-automatic deployment of at least one of the determined security policies to at least one security tool of the security tools requiring human intervention, or a simulated deployment of at least one of the determined security policies to at least one security tool of the security tools.

[0071] In some cases, the associated asset group comprises of assets that have been found to: (a) perform the given action, or (b) not perform the given action.

[0072] In some cases, the finding, based on the data, of one or more given assets that are associated with the given action can be directly from the data or indirectly by analyzing proxy data associated with other assets, other than the given assets.

[0073] In some cases, the obtained data is enriched from additional sources.

[0074] In some cases, the method further comprising: assessing, by the processing circuitry, the results of performing the action on the organization; and determining, by the processing circuitry, based on the assessment, one or more changes to the security policies.

[0075] In some cases, the prediction of productivity impact comprises quantitative assessment of one or more of:

[0076] (a) workflow interruptions measured by increases in task completion time;

[0077] (b) user experience degradation measured by: (i) additional authentication steps, (ii) access request frequency, or (iii) interface interactions, process delays resulting from security policy enforcement;

[0078] (c) resource utilization changes, including: CPU, memory, network, or storage consumption;

[0079] (d) application performance metrics, including: response time, throughput, or availability;

[0080] (e) collaboration friction introduced between one or more entities of the organization; (f) increased IT support requests;(g) increased security exception applications; or

[0081] (h) increased opportunity costs arising from restricted access to required resources.

[0082] In some cases, the method preforms at least part of the process repeatedly.

[0083] In some cases, the obtain includes obtaining new data associated with one or more of the assets of the organization and wherein the method further comprising: determining, by the processing circuitry, based on the new data, new security policies for one or more of the assets; detecting, by the processing circuitry, a security policy drift of one or more of the assets, being assets that have a difference above a threshold between the respective determined security policy and the respective new security policy; and automatically addressing, by the processing circuitry, the security policy drift by amending the respective determined security policy utilizing the new respective determined security policy.

[0084] In accordance with a third aspect of the presently disclosed subject matter, there is provided a non-transitory computer readable storage medium having computer readable program code embodied therewith, the computer readable program code, executable by at least one processing circuitry of a computer to perform a method for determination of one or more security policies, the method comprising: performing a process including: obtaining, by the processing circuitry: (A) data associated with one or more assets of an organization, and (B) information of one or more controls, each of the controls is associated with a security tool of one or more security tools of the organization, and is: (i) associated with a capability of a respective security tool of the security tools to prevent one or more given assets of the assets from taking one or more actions, and (ii) associated with one or more security policies of the security policies, wherein each security policy is a configuration of the given security tool that effects a group of assets of the assets; performing, by the processing circuitry, the following, for at least one given control of the controls, and for at least one given action of the actions associated with the given control: finding, by the processing circuitry, based on the data, one or more given assets that are associated with the given action, giving rise to an associated asset group; clustering, by the processing circuitry, the assets into one or more asset groups, wherein at least one asset group of the asset groups comprises one or more assets of the assets that have statistical similarity, and wherein the statistical similarity is determined by analyzing the associated asset group; and determining, by the processing circuitry, for at least one given asset group of the asset groups, a respective security policy, based on a predictionof a productivity impact of preventing the assets of the given asset group from preforming the given action; and performing, by the processing circuitry, an action based on the determined security policies.

[0085] BRIEF DESCRIPTION OF THE DRAWINGS

[0086] In order to understand the presently disclosed subject matter and to see how it may be carried out in practice, the subj ect matter will now be described, by way of non-limiting examples only, with reference to the accompanying drawings, in which:

[0087] Fig- 1 is a schematic illustration of an example of clustering one or more assets into asset groups, wherein at least one asset group of the asset groups comprises one or more assets of the assets that have statistical similarity, in accordance with the presently disclosed subject matter;

[0088] Fig- 2 is a block diagram schematically illustrating one example of a system for determining one or more security policies, in accordance with the presently disclosed subject matter; and

[0089] Fig- 3 is a flowchart illustrating an example of a sequence of operations carried out for determining one or more security policies, in accordance with the presently disclosed subject matter.

[0090] DETAILED DESCRIPTION

[0091] In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the presently disclosed subject matter. However, it will be understood by those skilled in the art that the presently disclosed subject matter may be practiced without these specific details. In other instances, well-known methods, procedures, and components have not been described in detail so as not to obscure the presently disclosed subject matter.

[0092] In the drawings and descriptions set forth, identical reference numerals indicate those components that are common to different embodiments or configurations.

[0093] Unless specifically stated otherwise, as apparent from the following discussions, it is appreciated that throughout the specification discussions utilizing terms such as "employing", "obtaining", "balancing", "identifying", "calculating", "generating", "performing", "clustering", "alerting", "determining" or the like, include action and / orprocesses of a computer that manipulate and / or transform data into other data, said data represented as physical quantities, e.g., such as electronic quantities, and / or said data representing the physical objects. The terms “computer”, “processor”, “processing resource”, “processing circuitry”, and “controller” should be expansively construed to cover any kind of electronic device with data processing capabilities, including, by way of non-limiting example, a personal desktop / laptop computer, a server, a computing system, a communication device, a smartphone, a tablet computer, a smart television, a processor (e.g. digital signal processor (DSP), a microcontroller, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), etc.), a group of multiple physical machines sharing performance of various tasks, virtual servers co-residing on a single physical machine, any other electronic computing device, and / or any combination thereof.

[0094] The operations in accordance with the teachings herein may be performed by a computer specially constructed for the desired purposes or by a general-purpose computer specially configured for the desired purpose by a computer program stored in a non-transitory computer readable storage medium. The term "non-transitory" is used herein to exclude transitory, propagating signals, but to otherwise include any volatile or nonvolatile computer memory technology suitable to the application.

[0095] As used herein, the phrase "for example”, "such as", "for instance" and variants thereof describe non-limiting embodiments of the presently disclosed subject matter. Reference in the specification to "one case", "some cases", "other cases" or variants thereof means that a particular feature, structure or characteristic described in connection with the embodiment s) is included in at least one embodiment of the presently disclosed subject matter. Thus, the appearance of the phrase "one case", "some cases", "other cases" or variants thereof does not necessarily refer to the same embodiment s).

[0096] As used herein, the phrase "one or more" is inclusive of "two or more”, "three or more”, or any other greater plurality. For instance, reference to "one or more" embodiments, features, or characteristics explicitly encompasses single embodiments, features, or characteristics as well as multiple embodiments, features, or characteristics thereby indicating no limitation regarding the exact number of embodiments, features, or characteristics covered. Thus, the use of "one or more" does not necessarily imply restriction to a specific number and is intended to broadly include any number equal to or greater than one.It is appreciated that, unless specifically stated otherwise, certain features of the presently disclosed subject matter, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the presently disclosed subject matter, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable sub-combination.

[0097] In embodiments of the presently disclosed subject matter, fewer, more and / or different stages than those shown in Fig. 3 may be executed. In embodiments of the presently disclosed subject matter one or more stages illustrated in Fig.3 may be executed in a different order and / or one or more groups of stages may be executed simultaneously.

[0098] Fig 2 illustrates a general schematic of the system architecture in accordance with an embodiment of the presently disclosed subject matter. Each module in Fig. 2 can be made up of any combination of software, hardware and / or firmware that performs the functions as defined and explained herein. The modules in Fig. 2 may be centralized in one location or dispersed over more than one location. In other embodiments of the presently disclosed subject matter, the system may comprise fewer, more, and / or different modules than those shown in Fig. 2.

[0099] Any reference in the specification to a method should be applied mutatis mutandis to a system capable of executing the method and should be applied mutatis mutandis to a non-transitory computer readable medium that stores instructions that once executed by a computer result in the execution of the method.

[0100] Any reference in the specification to a system should be applied mutatis mutandis to a method that may be executed by the system and should be applied mutatis mutandis to a non-transitory computer readable medium that stores instructions that may be executed by the system.

[0101] Any reference in the specification to a non-transitory computer readable medium should be applied mutatis mutandis to a system capable of executing the instructions stored in the non-transitory computer readable medium and should be applied mutatis mutandis to method that may be executed by a computer that reads the instructions stored in the non-transitory computer readable medium.

[0102] Attention is now drawn to Fig. 1, showing a non-limiting example of clustering one or more assets into asset groups, wherein at least one asset group of the asset groupscomprises one or more assets of the assets that have statistical similarity, in accordance with the presently disclosed subject matter.

[0103] As described above, the current organizational security policy landscape is characterized by increasing complexity due to the profusion of numerous security tools designed to address specific threats. Organizations (e.g., companies, non-profit organizations, governmental organizations, or any other type of organizations) today deploy firewalls, antivirus software, IDSs, identity and access management solutions, and cloud security platforms, among many other security tools.

[0104] Each of these security tools can implement one or more security controls. These controls are mechanisms designed to protect the organization by mitigating risks and preventing unauthorized access to physical premises, information systems, and other organizational assets. These controls include various settings and configurations that allow the security tools to control behavior. Examples include: Access Control List (ACL), IDSs, logging systems that track user access to sensitive assets, and many others. As an example, an ACL defines permissions attached to files or directories specifying which users or groups can access them and what operations (read, write, execute) they're permitted to perform. ACLs enforce precise access restrictions within the organization, preventing unauthorized users from viewing, modifying, or deleting sensitive information.

[0105] Each of these tools requires its own distinct security policies, configurations, and monitoring practices. Thus, organizations must continuously balance multiple, overlapping tools and increasingly sophisticated threats, all while managing employee productivity and satisfaction. As a result, the organization aspires to establish coherent, effective security policies across this fragmented tool landscape without negatively impacting productivity. A security policy includes, but is not limited to, security configuration, security settings, product configuration, product settings, and any other procedures established by the organization to ensure the protection of its assets. A security policy determines a specific configuration of a certain control of a security tool and to which assets is this configuration applicable.

[0106] A system for determination of one or more security policies creates organizationspecific security policies that are adjusted for the operational patterns and business needs of the organization while balancing maximum security, minimal productivity impact and maintaining Information Technology (IT) feasibility. IT feasibility is when the system fordetermination of one or more security policies evaluates whether specific security tools and security policies can be realistically and effectively implemented to achieve the desired security objectives within existing constraints of the organization's IT environment. For example, a firewall of an organization may have a technical limitation of no more than 1,000 different configurations. The system for determination of one or more security policies cannot determine configurations for the firewall that surpass its technical limitation. IT feasibility includes also maintainability of the security tools and their configurations. Maintainability is crucial, especially when dealing with large-scale security policy generation. For example, when automation systems could potentially create thousands of security policies. Excessive security policies pose a significant challenge for human administrators, who would find it impractical or even impossible to efficiently review, manage, and audit such extensive sets of security policies. Therefore, introducing intelligent logic designed to streamline and consolidate security policy creation is essential. The system for determination of one or more security policies utilizes a reduction approach to ensure the number of generated security policies is manageable without sacrificing security controls or hindering productivity. By embedding such logic within the system, policies remain effective yet practicable, enhancing both operational efficiency and overall security posture.

[0107] The system for determination of one or more security policies performs a process that includes obtaining data associated with one or more assets of the organization. It is noted that the phrase "one or more" is inclusive of "two or more”, "three or more”, or any other greater plurality. For instance, reference to "one or more" assets explicitly encompasses a single asset as well as multiple assets thereby indicating no limitation regarding the exact number of assets covered. Thus, the use of "one or more" does not necessarily imply restriction to a specific number and is intended to broadly include any number equal to or greater than one.

[0108] The organization can comprise a plurality of organizational assets (e.g., asset 110-1, asset 110-2, ..., asset 110-n). An asset can be a human (for example: an employee of the organization) or a non-human asset (for example: a device used by the employee). The organizational assets can include, but are not limited to, one or more of: a server, a computer, a laptop, a security tool, a network component, a smart device, a user, an application, a container, a virtual machine, a cloud service, an Internet of Things (loT) device, a mobile device, a database, an Application Programming Interface (API)endpoint, a software service, a data repository, data, a file, a document, a non-human identity, a service account, a machine identity, a digital certificate, a key, a token, a domain, a subnet, a network segment, or any other resource that is part of the organization, or that can connect to the organizational network or that is accessible via the organizational network. The assets can connect to an organizational network, or to parts thereof, via a wired and / or a wireless connection. As one can appreciate, the organizational assets can include users, Operational Technology (OT) devices, IT devices, or any other device of the organization. It is to be noted that the organizational network can be comprised of a plurality of sub-networks that can optionally be interconnected (whether unidirectionally or bidirectionally). It is to be noted that the term “asset” and the term “organizational asset” are used interchangeably throughout the detailed description.

[0109] The data obtained by the system for each given organizational asset can include logs of events associated with the given asset. Continuing our non-limiting example above, of the manufacturing company that needs to determine and implement application control security policies to prevent malware without disrupting critical engineering tools, an asset in this example can be a specialized CAD application. The data obtained by the system associated with the specialized CAD application can include logs of events performed by the specialized CAD application, for example: specific system calls that the specialized CAD application performed.

[0110] It is to be noted that the data obtained by the system for the assets can be in some cases deduced data - data that is derived from what the system learns for other assets. For example, assets that communicate with a given asset can inform the system on the behavior of the given asset - even when direct information of these communications is not available for the given asset. The data obtained by the system for each given asset can vary and also include, but not limited to, one or more of: organizational information of the given asset, information from external sources associated with the given asset, proxy data derived from one or more additional assets related to the given asset wherein direct visibility is limited, indirect logs from one or more interconnected systems, inferences drawn from absence of expected data, behavioral patterns identified through partial process observations, correlative data from sequential event analysis wherein observed initial steps without subsequent steps indicate incomplete execution chains, or synthesized visibility constructed from fragmented or incomplete monitoring data. Insome cases, the system can enrich the obtained data from additional sources. Some of these additional sources can be external to the system. For example, public data about specific workflow patterns and / or compliance requirements for organizations of certain business categories.

[0111] Continuing our non-limiting example above, the system can obtain information about the specialized CAD application from external sources. For example, that this type of specialized CAD application requires Internet connection to a specific IP address through a specific port. The system can also infer data associated with behavioral patterns of the specialized CAD application. For example, that the specialized CAD application performs periodic backups to an external drive once a week.

[0112] The system also obtains information of one or more controls. It is noted that the phrase "one or more" is inclusive of "two or more”, "three or more”, or any other greater plurality. For instance, reference to "one or more" controls explicitly encompasses a single control as well as multiple controls thereby indicating no limitation regarding the exact number of controls covered. Thus, the use of "one or more" does not necessarily imply restriction to a specific number and is intended to broadly include any number equal to or greater than one.

[0113] Each of these controls is associated with one of the security tools of the organization. Each control can be associated with a capability of the security tool to prevent one or more assets from taking an action (for example: accessing a certain file, communicating to a certain IP address through a specific port, etc.). Each control is also associated with one or more security policies. Each security policy is a configuration of the security tool that effects a given group of assets.

[0114] Continuing our non-limiting example above, the system can obtain information about controls of an exemplary security tool of the manufacturing company. For example, a firewall protecting the organizational network. The information includes IP address and which ports are open for communication and which addresses and ports are blocked.

[0115] The system for determining one or more security policies can now perform the following, for at least one given control of the controls, and for at least one given action of the actions associated with the given control: find, based on the obtained data, one or more given assets that are associated with the given action, giving rise to an associated asset group. The associated asset group can comprise of assets that have been found to perform the given action, or that the associated asset group can comprise of assets thathave been found not to perform the given action. These actions are performed in a nested loop, which is a loop that exists inside another loop. The outer loop runs first for at least some of the controls, and for every single iteration of the outer loop, the inner loop runs for at least one action associated with that control. In that way, at least some combinations of controls and tools are iterated by the system. Security policies are determined for each iterated combination. These are the determined security policies that the system can provide to a user of the system or perform other actions upon, as further detailed herein, inter alia with reference to Fig. 3.

[0116] Finding, based on the obtained data, of the one or more given assets that are associated with the given action can be performed directly from the obtained data or indirectly by analyzing proxy data associated with other assets, other than the given assets. For example, network traffic of a given asset may not be visible directly, but by analyzing organizational network data, the system reveals interactions between the given asset and other organizational assets. Similarly, if a relevant connection requires authentication, the system can infer that no connection was established if no authentication events are present in the corresponding logs. This indirect visibility approach allows the system to reasonably determine asset actions even with incomplete monitoring coverage.

[0117] Finding of the one or more given assets that are associated with the given action can be performed within a specific time window. The time windows can be set or continuously changing utilizing adaptive time windows, with or without relating to threat severity and asset criticality. Asset criticality can be determined, but is not limited to, utilizing the following approaches: security criticality, which assesses the potential security risk to the organization if the asset is compromised by an attacker, and / or productivity criticality, which evaluates the importance of the asset in maintaining business continuity. For example, assets that have performed a given action in the last 90 days will be part of the associated asset group for that given action. This window can be longer (for example, 180 days) for critical assets (for example, devices of the Chief Executive Officer (CEO) of the organization which are productivity critical and / or a firewall which is security critical).

[0118] It is noted that the system can decide not to allow a given action for one or more assets, even if these assets performed the given action in the past. For example, becausethe given action has been determined by the system to be a “one off’ (performed less than a threshold times in a given time window) for these assets.

[0119] Continuing our non-limiting example above, the system can find for the firewall’ s blocking ports control, for the blocking of a specific port, all the assets that have used that specific port in the last 90 days. These assets will form the associated asset group for that given action of blocking the specific port.

[0120] The system now clusters the assets into one or more asset groups (e.g., Asset group A 120-a, Asset group B 120-b, ..., Asset group M 120-m) in accordance with a statistical similarity that is based on the associated asset group. It is noted that the phrase "one or more" is inclusive of "two or more”, "three or more”, or any other greater plurality. For instance, reference to "one or more" asset groups explicitly encompasses a single asset group as well as multiple asset groups thereby indicating no limitation regarding the exact number of asset groups covered. Thus, the use of "one or more" does not necessarily imply restriction to a specific number and is intended to broadly include any number equal to or greater than one.

[0121] At least one asset group comprises one or more assets of the assets that have statistical similarity. Statistical similarity refers to the measure of closeness or resemblance between data points representing the assets based on their statistical characteristics or attributes. The statistical similarity can be determined by analyzing the associated asset group. The statistical similarity can be determined by the system once, or determined continuously by the system to adapt for changing asset behaviors and organizational requirements. In our non-limiting example depicted in Fig. 1, the system has clustered asset 110-2, asset 110-3, asset 110-5, asset 110-13, asset 110-15, asset 110-18, asset 110-19, asset 110-21 and asset 110-n into asset group A 120-a. This means that the system has found these assets to have statistical similarity. The statistical similarity is associated with the given action of the given control. In a similar way, the other assets depicted have been clustered into other asset groups (e.g., Asset group B 120-b, ... , Asset group M 120-m) based on their statistical similarity.

[0122] Continuing our non-limiting example above, the action is the usage of a specific port. The assets that have used the specific port in the last 90 days form the associated asset group. In our non-limiting example, the specific port is used to work with the specialized CAD application. All assets that worked with the specialized CAD application in the last 90 days will be in the associated asset group. The system utilizes the dataassociated with these assets to determine statistical similarity - data of the assets in the associated asset group is deemed statistically similar. The system then utilizes the determined statistical similarity to cluster all assets into asset groups in relation to the action of using the specific port. For example: one asset group comprises a high percentage of assets that used the specific port in the last 90 days (namely, the engineers that work with the specialized CAD application), another asset group comprises of a high percentage of assets that used the specific port in the last 90 days under a certain condition (namely, the managers that work with the specialized CAD application in view mode only) and a third asset group comprises of a high percentage of assets that did not use the specific port in the last 90 days (namely, other users of the organization that do not use the specialized CAD application).

[0123] In some cases, prior to the determination of the statistical similarity, the data associated with the assets of the performing asset group is pre-processed. For example, the pre-processing can be performed by removing one or more events associated with the given action. Continuing our non-limiting example above, the data associated with the assets of the performing asset group is pre-processed to remove log entries of the usage of the specific port to work with the specialized CAD application. This ensures that the clustering is not just based on the performance of the action itself but on other features of the clustered assets. For example, one of the engineers that usually works with the specialized CAD application has been on sick leave for the last 90 days. The asset representing that engineer would not be determined as part of the associated asset group by the system. The system will cluster this engineer with the other engineers that utilize the specialized CAD application because of other features in his data, features that are statistically similar to other engineers. This will allow the system to determine a security policy allowing the engineer that was on sick leave to work with the specialized CAD application even though he was not part of the performing asset group.

[0124] In some cases, the determination of the statistical similarity of the assets is preformed utilizing a machine learning model. A machine learning model is a computational system designed to learn patterns from data to make predictions or decisions without explicit programming. These models use statistical methods and modern computing to analyze large datasets and automate the discovery of insights. Once trained, machine learning models are deployed in diverse applications like imagerecognition, natural language processing, and fraud detection, often outperforming traditional methods.

[0125] Machine learning models fall into two main categories: supervised and unsupervised learning. Supervised learning involves training on labeled data, where the model learns to predict labels for new data. Examples include Linear Regression for continuous predictions, Logistic Regression for binary classifications, Decision Trees, Random Forests, XGBoost, Support Vector Machines (SVM), and Neural Networks, effective for complex tasks such as image and speech recognition.

[0126] Unsupervised learning models analyze unlabeled data to detect inherent patterns, groupings, or features. Common algorithms include K-Means Clustering, DBSCAN, and Principal Component Analysis (PCA), which reduces data dimensions while preserving key information. Clustering, involves grouping data points into clusters based on statistical similarity measures. Algorithms such as K-Means partition data into distinct groups by minimizing the distance to cluster centroids, whereas DBSCAN identifies clusters based on density, allowing it to handle arbitrary shapes and detect outliers. Clustering helps uncover hidden structures within data, enabling exploratory analysis, customer segmentation, anomaly detection, and pattern recognition. In machine learning and clustering, statistical similarity determines how data points are grouped together or separated. Two points with high statistical similarity share common patterns or feature values, indicating they likely belong to the same cluster or category. Conversely, points with low similarity differ significantly, suggesting they belong to distinct groups. The machine learning model utilized by the system can be a machine learning model that is capable of receiving the pre-processed data associated with a given asset of the assets, and determining an embedding vector representative of the asset. The machine learning model can be trained utilizing the pre-processed data associated with the assets of the performing asset group. In this case, the clustering of the assets into one or more asset groups is preformed utilizing the embedding vectors representative of the respective assets.

[0127] Continuing our non-limiting example above, the data associated with the assets of the performing asset group is used to train the machine learning model that produces embedding vectors for data associated with an asset. The system utilizes the machine learning model to generate embedding vectors for all the assets. Each vector represents an asset. The system clusters the vectors into asset groups. The vector representative ofthe engineer that was on sick leave will be clustered by the system with the vectors representative of the other engineers that utilize the specialized CAD application.

[0128] The system can then determine, for at least one given asset group of the asset groups, a respective security policy, based on a prediction of a productivity impact of preventing the assets of the given asset group from preforming the given action. The system can utilize the determined security policies to perform an action, as further detailed herein, inter alia with reference to Fig. 3.

[0129] After describing a non-limiting example of clustering one or more assets into asset groups, wherein at least one asset group of the asset groups comprises one or more assets of the assets that have statistical similarity, attention is now drawn to Fig. 2.

[0130] Fig.2, depicts a block diagram schematically illustrating one example of a system for determining one or more security policies for an organization, in accordance with the presently disclosed subject matter.

[0131] In accordance with the presently disclosed subject matter, the system for determining one or more security policies (also referred herein as: “system 200”) can comprise a network interface 220. The network interface 220 (e.g., a network card, a WiFi client, a Li-Fi client, 3G / 4G / 5G client, satellite communications or any other component), enables system 200 to communicate over a network with external systems and handles inbound and outbound communications from such systems. For example, system 200 can receive and / or send, through network interface 220, data associated with one or more assets of the organization, information of one or more security tools of the organization, information of one or more controls associated with the security tools, settings and / or configurations associated with the controls, security policies, etc.

[0132] System 200 can be part of the organizational network and / or external to the organizational network. In some cases, system 200 or parts thereof reside on a cloud computing environment.

[0133] System 200 can further comprise or be otherwise associated with a data repository 210 (e.g., a database, a storage system, a blockchain based storage, a memory including Read Only Memory - ROM, Random Access Memory - RAM, or any other type of memory and / or storage, etc.) configured to store data. Some examples of data that can be stored in the data repository 210 include: data associated with one or more assets (e.g., asset 110-1, asset 110-2, ..., asset 110-n) of the organization, information of one or more security tools of the organization, information of one or more controls associated with thesecurity tools, settings and / or configurations associated with the controls, security policies, etc. Data repository 210 can be further configured to enable retrieval and / or update and / or deletion of the stored data. It is to be noted that in some cases, data repository 210 can be distributed, and / or reside at least partly on a cloud computing environment, while system 200 has access to the information stored thereon, e.g., via a wired or wireless network to which system 200 is able to connect (utilizing its network interface 220).

[0134] System 200 further comprises processing circuitry 230. Processing circuitry 230 can be one or more processing units (e.g., central processing units), microprocessors, microcontrollers (e.g., microcontroller units (MCUs) cloud servers, graphical processing units (GPUs), or any other computing devices or modules, including multiple and / or parallel and / or distributed processing units, which are adapted to independently or cooperatively process data for controlling relevant system 200 resources and for enabling operations related to system’s 200 resources.

[0135] The processing circuitry 230 comprises a security policies determination module 240, configured to perform a process for determining one or more security policies, as further detailed herein, inter alia with reference to Fig. 3.

[0136] Having described the block diagram of system 200, attention is now drawn to Fig.

[0137] 3 showing a flowchart illustrating an example of a sequence of operations carried out for determining one or more security policies, in accordance with the presently disclosed subject matter. Accordingly, system 200 can be configured to perform a process for determining one or more security policies 300, e.g., using the security policies determination module 240. The process includes the following:

[0138] System 200 first obtains: (A) data associated with one or more assets (e.g., asset 110-1, asset 110-2, ..., asset 110-n) of an organization, and (B) information of one or more controls, each of the controls is associated with a security tool of one or more security tools of the organization, and is: (i) associated with a capability of a respective security tool of the security tools to prevent one or more given assets of the assets from taking one or more actions, and (ii) associated with one or more security policies of the security policies, wherein each security policy is a configuration of the given security tool that effects a group of assets of the assets (block 310).

[0139] Continuing our non-limiting example above, of the manufacturing company that needs to implement application control security policies to prevent malware withoutdisrupting critical engineering tools. In this example, system 200 obtains data associated with the assets of the manufacturing company. This data can include: which applications engineers use, when, and how they interact with system resources. The data can be enriched with organizational context and / or public context, like: engineering department structure, critical manufacturing systems, and relevant regulation. The data can also be obtained by system 200 from external sources, like: knowledge about how engineering tools typically function across the manufacturing sector. System 200 also obtains information about the controls of the security tools of the manufacturing company. For example, port blocking controls of the firewall used by the company.

[0140] After obtaining the data associated with the one or more assets and the information of the one or more controls, system 200 can perform the following, for at least one given control of the controls, and for at least one given action of the actions associated with the given control: find, based on the data, one or more given assets that are associated with the given action, giving rise to an associated asset group; cluster the assets into one or more asset groups, wherein at least one asset group of the asset groups comprises one or more assets of the assets that have statistical similarity, and wherein the statistical similarity is determined by analyzing the associated asset group; and determine, for at least one given asset group of the asset groups, a respective security policy, based on a prediction of a productivity impact of preventing the assets of the given asset group from preforming the given action (block 320).

[0141] As detailed above, inter alia with reference to Fig. 1, for at least one given control of the controls, and for at least one given action of the actions associated with the given control, system 200 clusters the assets into one or more asset groups (e.g., Asset group A 120-a, Asset group B 120-b, ..., Asset group M 120-m) in accordance with a statistical similarity that is associated with the given action. As the clustering is performed for at least one action (and can be performed for each of the actions) that are associated with at least one control (and can be performed for each of the controls). Thus, system 200 has multiple groupings of the assets into clusters. Each grouping is associated with a statistical similarity that is associated with the given action.

[0142] After clustering the assets into asset groups, system 200 determines, for at least one given asset group of the asset groups, a respective security policy. The determination of the security policy for the given asset group is based on a prediction of a productivity impact of preventing the assets of the given asset group from preforming the given action.System 200 can determine the respective security policy by utilizing one or more rules. The rules are used to determine, for at least one given asset group of the asset groups, a respective security policy, based on a prediction of a productivity impact of preventing the assets of the given asset group from preforming the given action. It is noted that the phrase "one or more" is inclusive of "two or more", "three or more", or any other greater plurality. For instance, reference to "one or more" rules explicitly encompasses a single rule as well as multiple rules thereby indicating no limitation regarding the exact number of rules covered. Thus, the use of "one or more" does not necessarily imply restriction to a specific number and is intended to broadly include any number equal to or greater than one.

[0143] Continuing our non-limiting example above wherein the action is the usage of a specific port. This action is part of a control that controls ports as part of the firewall. System 200 in this example clustered the assets into the following asset groups: asset group A 120-a comprises a high percentage of assets that used the specific port in the last 90 days (namely, the engineers that work with the specialized CAD application), asset group B 120-b comprises of a high percentage of assets that used the specific port in the last 90 days under a certain condition (namely, the managers that work with the specialized CAD application in view mode only) and asset group M 120-m comprises of a high percentage of assets that did not utilize the specific port (for example, by receiving a connection to the port and / or sending a connection to the port) in the last 90 days (namely, other users of the organization that do not use the specialized CAD application). After clustering the assets into the asset groups, system 200 determines a certain security policy associated with the given action for each of the asset groups. This certain security policy applies to all assets in that asset group. In our example, the rule can be that if 90% or more of the assets in asset group A 120-a used the specific port in the last 90 days -the policy for the assets in asset group A 120-a for the action of the usage of a specific port is that they are allowed to use the specific port. That rule can be that if 90% or more of the assets in asset group B 120-b used the specific port in the last 90 days under a certain condition - the policy for the assets in asset group B 120-b for the action of the usage of a specific port is that they are allowed to use the specific port, but only under the condition. That rule can be that if 90% or more of the assets in asset group C 120-c did not use the specific port in the last 90 days - the policy for the assets in asset group C 120-c for the action of the usage of a specific port is that they are not allowed to use thespecific port. It is to be noted that this security policy as determined by system 200 will allow the engineer that was on sick leave for the last 90 days, from the example above, to use the specific port because system 200 has determined statistical similarity between him and the other engineers and has clustered him into the same asset group as the other engineers.

[0144] System 200 determines which rule to activate on the asset groups in accordance with a threshold that is associated with an importance of the security policy. The importance of the security policy can change in accordance with the type of organization. The rules can also balance a calculated productivity impact on the assets of the given asset group and the security needs of the organization.

[0145] Continuing our non-limiting example above, the rule that examines if 90% or more of the assets in the asset group A 120-a has performed the action balances the productivity impact by allowing these assets to use the specific port thereby supporting their workflow of using the specialized CAD application.

[0146] The prediction of productivity impact can also include quantitative assessment of one or more of

[0147] (a) workflow interruptions measured by increases in task completion time.

[0148] (b) user experience degradation measured by: (i) additional authentication steps, (ii) access request frequency, or (iii) interface interactions, process delays resulting from security policy enforcement.

[0149] (c) resource utilization changes, including: CPU, memory, network, or storage consumption.

[0150] (d) application performance metrics, including: response time, throughput, or availability.

[0151] (e) collaboration friction introduced between one or more entities of the organization. (f) increased IT support requests.

[0152] (g) increased security exception applications.

[0153] (h) increased opportunity costs arising from restricted access to required resources.

[0154] System 200 can utilize the rules in a way that reflects a given level of allowed false positives. For example, in an organization with 1000 assets, the system is allowed 100 false positives within a given timeframe. The allowed false positives are distributed between the assets in accordance with organizational information of the respective assets, between the controls in accordance with organizational information of the respectivecontrols, or between the actions in accordance with organizational information of the respective actions.

[0155] Each security tool has technical limitation with regards with the number of security policies and / or the types of security policies supported by the security tool. For example, a port control of a firewall may support only a limited number of security policies. If the determination of the security policies by system 200 has generated a number of security policies for the control that surpasses the technical capability of the security tool, system 200 apply only part of the security policies and / or merge the security policies of the security tool in accordance with the IT feasibility. In some cases, system 200 can select part of the determined security policies into a new security policy. This new security policy is partial, but is within the technical limitations of the security tool.

[0156] System 200 can also split or merge at least two of the determined security policies associated with the respective given control into a new security policy which is an optimized security policy with regards to the technical limitations of the security tool. System 200 will utilize the new security policy when performing the action, as detailed below in block 330. The merging or splitting of the determined security policies into a new optimized security policy can be performed utilizing a heuristic technique.

[0157] Continuing our non-limiting example above, system 200 has determined a policy for each control of the firewall in accordance with the clustering of the assets into asset groups. This results in a separate port control policy for each asset. System 200 can merge the assets that have common policies together into one security policy, but this may not be sufficient as the number of unique port policies can still exceed the technical limitation of the firewall. System 200 utilizes a heuristic approach to merge the policies to a number that is feasible for the firewall. The heuristic approach balances productivity and security. The heuristic approach can be based on a security value associated with the respective action. For example, if the action is for a specific port that has a high security value than the heuristic approach will be more on the security side and will allow less productivity when merging the security policies into the optimized security policy.

[0158] The security values is a resilience score of the action. The security score can be determined based on one or more Tactics, Techniques and Procedures (TTPs) that involve the given action, for example based on the techniques of the MITRE ATTACK.

[0159] Continuing our non-limiting example above, system 200 can determine a high score for the action for a specific port when this port is used repeatedly in MITREATTACK techniques. A port that is less used in MITRE ATTACK techniques will receive a lower security value.

[0160] In some cases, the TTPs can be weighted in accordance with one or more characteristics of the organization. In other cases, the TTPs are weighted in accordance with a risk level. For example, our exemplary organization is a manufacturing company than certain MITRE ATTACK techniques are more relevant and will receive a higher wight than other MITRE ATTACK techniques when determining the security value of the action. Similarly, the security value can be weighted in accordance with a risk associated with each MITRE ATTACK technique.

[0161] After determine the one or more security policies and, if required, optimizing them in accordance with the technical limitations of the security tools, system 200 can perform an action based on the determined one or more security policies (block 330).

[0162] In some cases, the action is a recommendation to a user of the system 200 to deploy the security policies. In other cases, system 200 can automatically deploy at least one of the determined security policies to at least one security tool of the security tools. Another possibility for the action is a semi-automatic deployment of at least one of the determined security policies to at least one security tool of the security tools requiring human intervention. In some cases, that action is a simulated deployment of at least one of the determined security policies to at least one security tool of the security tools.

[0163] Continuing the non-limiting example above, system 200 has determined that certain ports of the firewall should be open for some of the assets and that other ports should be blocked for other assets. System 200 can recommend these policies to a user of the system 200. System 200 can simulate the impact of deploying these security policies to the firewall and depict the results of the simulation to the user, thereby allowing them to decide if and how to deploy the security policies. The system 200 can also support deploying the security policies to the security tools either fully automatically or semi-automatically - requiring a human approval during the deployment.

[0164] The simulation can include a prediction of an operational impact of the simulated deployment on one or more assets of the assets, thereby allowing the user to understand the impact of the deployment of the security policies before actually impacting the organization.

[0165] Continuing the non-limiting example above, if system 200 has determined a security policy that includes blocking of the specific port for the CEO of the company, hewill not be able to use the specialized CAD application. A simulation will predict this impact and represent it to the user of system 200. The user may decide to deploy the security policies in a way that excludes the blocking of the CEO.

[0166] The simulation can include a prediction of an expected risk reduction of the simulated deployment on one or more assets of the assets. The expected risk reduction can be quantified across one or more threat vectors and / or attack scenarios, thereby allowing the user to understand the impact of the deployment of the security policies before actually impacting the organization.

[0167] Continuing the non-limiting example above, if system 200 has determined a security policy that includes blocking of the specific port for some of the users of the company and the specific port is part of a certain attack scenario, then the simulation will quantify the risk reduction in the possible occurrence of this certain attack scenario and represent it to the user of system 200.

[0168] In some cases, system 200 can assess the results of performing the action on the organization before performing the action and determine, based on the assessment, one or more changes to the security policies. These changes can include: monitoring actual productivity impacts following implementation of the determined security policy, comparing the actual productivity impacts with the predicted productivity impacts, adjusting predictive models based on discrepancies between actual and predicted impacts, automatically refining subsequent security policy determinations using the adjusted predictive models, thereby creating a continuous improvement cycle that increases prediction accuracy over time.

[0169] Continuing the non-limiting example above, system 200 can assess the results of blocking the specific port on one or more assets of the organization that require the usage of the specialized CAD application and change the security policy to exclude these specific assets from the security policies. The action will then be performed by system 200 without impacting these assets.

[0170] It is to be noted, that the herein-described process for determining one or more security policies 300 can be performed, at least partly, repeatedly and / or continually and / or periodically by system 200.

[0171] Continuing the non-limiting example above, system 200 has deployed the determined security policies at one point in time. System 200 can re-iterate process 300after a certain time period to determine and deploy new security policies based on freshly obtained data and information about the assets and the controls of the organization.

[0172] In some cases, system 200 can repeatedly perform process 300 to detect a security policy drift of one or more of the assets. These are assets that have a difference above a threshold between a security policy determined in the current run of the process 300 and the respective security policy of the previous run of process 300. System 200 can automatically address the security policy drift by amending the respective.

[0173] Continuing the non -limiting example above, system 200 runs process 300 on the organization to obtain new events that have occurred since the last time data and information has been retrieved. This process brings in new behavioral insights, as well as organizational data that may have been updated in the interim. System 200 combines this fresh information with the previously collected data and analyze them together. Using this consolidated dataset, system 200 performs calculating similarities, predicting potential productivity impact, computing optimal policies, and merging these into actionable security policies. System 200 optimizes the security policies according to the organization's needs - such as: focusing on the threats most relevant to the organization, statistics about common attack methods within these types of organizations, and global trends. System 200 incorporates the latest obtained data into the newly generated security policies that are more fitted to the current state of the organization. In our example, an asset of a newly recruited engineer can be added in the new run of process 300 to the security policy that allows them to use the specific port needed for the specialized CAD application.

[0174] A security policy drift of one or more of the assets is indicatory of a gap between an ideal state represented by newly calculated security policies and the current state which are the security policies determined by system 200 in the past and are currently implemented within the organization. Security policy drift can be caused by changes in organizational behavior, shifts in threat priorities or interests within the organization, updates in threat statistics or weighting within our algorithms, addition of new assets into the organization's systems or any other internal or external change.

[0175] As the security policy drift increases over time, so does a potential benefit to security and productivity, should the user of system 200 decide to adopt the updated determined security policies for a given control. The security policy drift forms the basis for a prioritization mechanism: for each change in the security policy, system 200 canassess a potential gain the organization will achieve by replacing the existing security policies with the newly determined security policies. System 200 can automatically, or semi-automatically, deploy the new security policies when the gain is above a threshold.

[0176] It is to be noted, with reference to Fig.3, that some of the blocks can be integrated into a consolidated block or can be broken down to a few blocks and / or other blocks may be added. It is to be further noted that some of the blocks are optional. It should be also noted that whilst the flow diagram is described also with reference to the system elements that realizes them, this is by no means binding, and the blocks can be performed by elements other than those described herein.

[0177] It is to be understood that the presently disclosed subject matter is not limited in its application to the details set forth in the description contained herein or illustrated in the drawings. The presently disclosed subject matter is capable of other embodiments and of being practiced and carried out in various ways. Hence, it is to be understood that the phraseology and terminology employed herein are for the purpose of description and should not be regarded as limiting. As such, those skilled in the art will appreciate that the conception upon which this disclosure is based may readily be utilized as a basis for designing other structures, methods, and systems for carrying out the several purposes of the present presently disclosed subject matter.

[0178] It will also be understood that the system according to the presently disclosed subject matter can be implemented, at least partly, as a suitably programmed computer. Likewise, the presently disclosed subject matter contemplates a computer program being readable by a computer for executing the disclosed method. The presently disclosed subject matter further contemplates a machine-readable memory tangibly embodying a program of instructions executable by the machine for executing the disclosed method.

Claims

CLAIMS:

1. A system for determination of one or more security policies, the system comprising a processing circuitry configured to perform a process including: obtain:(A) data associated with one or more assets of an organization, and(B) information of one or more controls, each of the controls is associated with a security tool of one or more security tools of the organization, and is: (i) associated with a capability of a respective security tool of the security tools to prevent one or more given assets of the assets from taking one or more actions, and (ii) associated with one or more security policies of the security policies, wherein each security policy is a configuration of the given security tool that effects a group of assets of the assets;perform the following, for at least one given control of the controls, and for at least one given action of the actions associated with the given control:find, based on the data, one or more given assets that are associated with the given action, giving rise to an associated asset group;cluster the assets into one or more asset groups, wherein at least one asset group of the asset groups comprises one or more assets of the assets that have statistical similarity, and wherein the statistical similarity is determined by analyzing the associated asset group; anddetermine, for at least one given asset group of the asset groups, a respective security policy, based on a prediction of a productivity impact of preventing the assets of the given asset group from preforming the given action; andperform an action based on the determined security policies.

2. The system of claim 1, wherein the processing circuitry is further configured to:apply a given security policy of the security policies on at least one given control of the controls, based on an Information Technology (IT) feasibility of the security tool associated with the given control.

3. The system of claim 2, wherein the apply includes merging at least two of the determined security policies associated with the respective given control into a mergeddetermined security policy and wherein the action is performed on the merged determined security policies.

4. The system of claim 3, wherein the merging of the at least two of the determined security policies into a merged security policy is based on a security value of the action associated with the respective security policy.

5. The system of claim 4, wherein the security value of a given action is determined based on one or more Tactics, Techniques and Procedures (TTPs) that involve the given action.

6. The system of claim 5, wherein the TTPs are based on MITRE.

7. The system of claim 5, wherein the TTPs are weighted in accordance with one or more characteristics of the organization.

8. The system of claim 5, wherein the TTPs are weighted in accordance with a risk level.

9. The system of claim 3, wherein the merging of the at least two of the determined security policies into a merged security policy is performed utilizing a heuristic technique.

10. The system of claim 2, wherein the apply includes selecting at least one of the determined security policies associated with the respective given control into a selected determined security policy and wherein the action is performed on the selected determined security policies.

11. The system of claim 1, wherein prior to the determination of the statistical similarity, the data associated with the assets of the performing asset group is pre-processed.

12. The system of claim 11, wherein the pre-processing is performed by removing one or more events associated with the given action.

13. The system of claim 11, wherein the determination of the statistical similarity of the assets is preformed utilizing a machine learning model.

14. The system of claim 13, wherein the machine learning model is capable of receiving: the pre-processed data associated with a given asset of the assets, and determining an embedding vector representative of the asset, wherein the machine learning model is trained utilizing the pre-processed data associated with the assets of the performing asset group.

15. The system of claim 14, wherein the clustering of the assets into one or more asset groups is preformed utilizing the embedding vectors representative of the respective assets.

16. The system of claim 1, wherein the determination of the respective security policy is preformed utilizing one or more rules balancing a calculated productivity impact on the assets of the given asset group and security needs of the organization.

17. The system of claim 16, wherein the rules reflect a given level of allowed false positives.

18. The system of claim 17, wherein the allowed false positives are distributed between one or more of: (a) the assets in accordance with organizational information of the respective assets, (b) the controls in accordance with organizational information of the respective controls, or (c) the actions in accordance with organizational information of the respective actions.

19. The system of claim 1, wherein the finding of the one or more given assets that are associated with the given action is performed within a time window.

20. The system of claim 1, wherein the one or more assets can include one or more of: a server, a computer, a laptop, a security tool, a network component, a smart device, a user, an application, a container, a virtual machine, a cloud service, an loT device, amobile device, a database, an API endpoint, a software service, a data repository, data, a file, a document, a non-human identity, a service account, a machine identity, a digital certificate, a key, a token, a domain, a subnet, or a network segment.

21. The system of claim 1, wherein the obtained data can include one or more of: logs of events associated with the respective asset, organizational information of the respective asset, information from external sources associated with the respective asset, proxy data derived from one or more additional assets related to the respective asset wherein direct visibility is limited, indirect logs from one or more interconnected systems, inferences drawn from absence of expected data, behavioral patterns identified through partial process observations, correlative data from sequential event analysis wherein observed initial steps without subsequent steps indicate incomplete execution chains, or synthesized visibility constructed from fragmented or incomplete monitoring data.

22. The system of claim 1, wherein the processing circuitry is further configured to:simulate, before performing the action, a deployment of at least one of the determined security policies to at least one security tool of the security tools; and predict an operational impact of the simulated deployment on one or more assets of the assets.

23. The system of claim 1, wherein the processing circuitry is further configured to:simulate, before performing the action, a deployment of at least one of the determined security policies to at least one security tool of the security tools; and quantify an expected risk reduction of the simulated deployment on one or more assets of the assets.

24. The system of claim 1, wherein the action is one or more of: a recommendation to a user of the system, an automatic deployment of at least one of the determined security policies to at least one security tool of the security tools, a semi-automatic deployment of at least one of the determined security policies to at least one security tool of the security tools requiring human intervention, or a simulated deployment of at least one of the determined security policies to at least one security tool of the security tools.

25. The system of claim 1, wherein the associated asset group comprises of assets that have been found to: (a) perform the given action, or (b) not perform the given action.

26. The system of claim 1, wherein the finding, based on the data, of one or more given assets that are associated with the given action can be directly from the data or indirectly by analyzing proxy data associated with other assets, other than the given assets.

27. The system of claim 1, wherein the obtained data is enriched from additional sources.

28. The system of claim 1, wherein the processing circuitry is further configured to:asses the results of performing the action on the organization; and determine, based on the assessment, one or more changes to the security policies.

29. The system of claim 1, wherein the prediction of productivity impact comprises quantitative assessment of one or more of:(a) workflow interruptions measured by increases in task completion time;(b) user experience degradation measured by: (i) additional authentication steps, (ii) access request frequency, or (iii) interface interactions, process delays resulting from security policy enforcement;(c) resource utilization changes, including: CPU, memory, network, or storage consumption;(d) application performance metrics, including: response time, throughput, or availability;(e) collaboration friction introduced between one or more entities of the organization; (f) increased IT support requests;(g) increased security exception applications; or(h) increased opportunity costs arising from restricted access to required resources.

30. The system of claim 1, wherein the processing circuitry preforms at least part of the process repeatedly.

31. The system of claim 30, wherein the obtain includes obtaining new data associated with one or more of the assets of the organization and wherein the processing circuitry is further configured to:determine, based on the new data, new security policies for one or more of the assets;detect a security policy drift of one or more of the assets, being assets that have a difference above a threshold between the respective determined security policy and the respective new security policy; andautomatically address the security policy drift by amending the respective determined security policy utilizing the new respective determined security policy.

32. A method for determination of one or more security policies, the method comprising: performing a process including:obtaining, by a processing circuitry:(A) data associated with one or more assets of an organization, and(B) information of one or more controls, each of the controls is associated with a security tool of one or more security tools of the organization, and is: (i) associated with a capability of a respective security tool of the security tools to prevent one or more given assets of the assets from taking one or more actions, and (ii) associated with one or more security policies of the security policies, wherein each security policy is a configuration of the given security tool that effects a group of assets of the assets;performing, by the processing circuitry, the following, for at least one given control of the controls, and for at least one given action of the actions associated with the given control:finding, by the processing circuitry, based on the data, one or more given assets that are associated with the given action, giving rise to an associated asset group;clustering, by the processing circuitry, the assets into one or more asset groups, wherein at least one asset group of the asset groups comprises one or more assets of the assets that have statistical similarity, and wherein the statistical similarity is determined by analyzing the associated asset group; anddetermining, by the processing circuitry, for at least one given asset group of the asset groups, a respective security policy, based on a prediction of a productivity impact of preventing the assets of the given asset group from preforming the given action; andperforming, by the processing circuitry, an action based on the determined security policies.

33. The method of claim 32, wherein the method further comprising:applying, by the processing circuitry, a given security policy of the security policies on at least one given control of the controls, based on an Information Technology (IT) feasibility of the security tool associated with the given control.

34. The method of claim 33, wherein the apply includes merging at least two of the determined security policies associated with the respective given control into a merged determined security policy and wherein the action is performed on the merged determined security policies.

35. The method of claim 34, wherein the merging of the at least two of the determined security policies into a merged security policy is based on a security value of the action associated with the respective security policy.

36. The method of claim 35, wherein the security value of a given action is determined based on one or more Tactics, Techniques and Procedures (TTPs) that involve the given action.

37. The method of claim 36, wherein the TTPs are based on MITRE.

38. The method of claim 36, wherein the TTPs are weighted in accordance with one or more characteristics of the organization.

39. The method of claim 36, wherein the TTPs are weighted in accordance with a risk level.

40. The method of claim 34, wherein the merging of the at least two of the determined security policies into a merged security policy is performed utilizing a heuristic technique.

41. The method of claim 33, wherein the apply includes selecting at least one of the determined security policies associated with the respective given control into a selected determined security policy and wherein the action is performed on the selected determined security policies.

42. The method of claim 32, wherein prior to the determination of the statistical similarity, the data associated with the assets of the performing asset group is pre-processed.

43. The method of claim 42, wherein the pre-processing is performed by removing one or more events associated with the given action.

44. The method of claim 42, wherein the determination of the statistical similarity of the assets is preformed utilizing a machine learning model.

45. The method of claim 44, wherein the machine learning model is capable of receiving: the pre-processed data associated with a given asset of the assets, and determining an embedding vector representative of the asset, wherein the machine learning model is trained utilizing the pre-processed data associated with the assets of the performing asset group.

46. The method of claim 45, wherein the clustering of the assets into one or more asset groups is preformed utilizing the embedding vectors representative of the respective assets.

47. The method of claim 32, wherein the determination of the respective security policy is preformed utilizing one or more rules balancing a calculated productivity impact on the assets of the given asset group and security needs of the organization.

48. The method of claim 47, wherein the rules reflect a given level of allowed false positives.

49. The method of claim 48, wherein the allowed false positives are distributed between one or more of: (a) the assets in accordance with organizational information of the respective assets, (b) the controls in accordance with organizational information of the respective controls, or (c) the actions in accordance with organizational information of the respective actions.

50. The method of claim 32, wherein the finding of the one or more given assets that are associated with the given action is performed within a time window.

51. The method of claim 32, wherein the one or more assets can include one or more of: a server, a computer, a laptop, a security tool, a network component, a smart device, a user, an application, a container, a virtual machine, a cloud service, an loT device, a mobile device, a database, an API endpoint, a software service, a data repository, data, a file, a document, a non-human identity, a service account, a machine identity, a digital certificate, a key, a token, a domain, a subnet, or a network segment.

52. The method of claim 32, wherein the obtained data can include one or more of: logs of events associated with the respective asset, organizational information of the respective asset, information from external sources associated with the respective asset, proxy data derived from one or more additional assets related to the respective asset wherein direct visibility is limited, indirect logs from one or more interconnected systems, inferences drawn from absence of expected data, behavioral patterns identified through partial process observations, correlative data from sequential event analysis wherein observed initial steps without subsequent steps indicate incomplete execution chains, or synthesized visibility constructed from fragmented or incomplete monitoring data.

53. The method of claim 32, wherein the method further comprising:simulating, by the processing circuitry, before performing the action, a deployment of at least one of the determined security policies to at least one security tool of the security tools; andpredicting, by the processing circuitry, an operational impact of the simulated deployment on one or more assets of the assets.

54. The method of claim 32, wherein the method further comprising: simulating, by the processing circuitry, before performing the action, a deployment of at least one of the determined security policies to at least one security tool of the security tools; andquantifying, by the processing circuitry, an expected risk reduction of the simulated deployment on one or more assets of the assets.

55. The method of claim 32, wherein the action is one or more of: a recommendation to a user of the system, an automatic deployment of at least one of the determined security policies to at least one security tool of the security tools, a semi-automatic deployment of at least one of the determined security policies to at least one security tool of the security tools requiring human intervention, or a simulated deployment of at least one of the determined security policies to at least one security tool of the security tools.

56. The method of claim 32, wherein the associated asset group comprises of assets that have been found to: (a) perform the given action, or (b) not perform the given action.

57. The method of claim 32, wherein the finding, based on the data, of one or more given assets that are associated with the given action can be directly from the data or indirectly by analyzing proxy data associated with other assets, other than the given assets.

58. The method of claim 32, wherein the obtained data is enriched from additional sources.

59. The method of claim 32, wherein the method further comprising:assessing, by the processing circuitry, the results of performing the action on the organization; anddetermining, by the processing circuitry, based on the assessment, one or more changes to the security policies.

60. The method of claim 32, wherein the prediction of productivity impact comprises quantitative assessment of one or more of:(a) workflow interruptions measured by increases in task completion time;(b) user experience degradation measured by: (i) additional authentication steps, (ii) access request frequency, or (iii) interface interactions, process delays resulting from security policy enforcement;(c) resource utilization changes, including: CPU, memory, network, or storage consumption;(d) application performance metrics, including: response time, throughput, or availability;(e) collaboration friction introduced between one or more entities of the organization; (f) increased IT support requests;(g) increased security exception applications; or(h) increased opportunity costs arising from restricted access to required resources.

61. The method of claim 32, wherein the method preforms at least part of the process repeatedly.

62. The method of claim 61, wherein the obtain includes obtaining new data associated with one or more of the assets of the organization and wherein the method further comprising:determining, by the processing circuitry, based on the new data, new security policies for one or more of the assets;detecting, by the processing circuitry, a security policy drift of one or more of the assets, being assets that have a difference above a threshold between the respective determined security policy and the respective new security policy; and automatically addressing, by the processing circuitry, the security policy drift by amending the respective determined security policy utilizing the new respective determined security policy.

63. A non-transitory computer readable storage medium having computer readable program code embodied therewith, the computer readable program code, executable by at least one processing circuitry of a computer to perform a method for determination of one or more security policies, the method comprising: performing a process including:obtaining, by the processing circuitry:(A) data associated with one or more assets of an organization, and(B) information of one or more controls, each of the controls is associated with a security tool of one or more security tools of the organization, and is: (i) associated with a capability of a respective security tool of the security tools to prevent one or more given assets of the assets from taking one or more actions, and (ii) associated with one or more security policies of the security policies, wherein each security policy is a configuration of the given security tool that effects a group of assets of the assets;performing, by the processing circuitry, the following, for at least one given control of the controls, and for at least one given action of the actions associated with the given control:finding, by the processing circuitry, based on the data, one or more given assets that are associated with the given action, giving rise to an associated asset group;clustering, by the processing circuitry, the assets into one or more asset groups, wherein at least one asset group of the asset groups comprises one or more assets of the assets that have statistical similarity, and wherein the statistical similarity is determined by analyzing the associated asset group; anddetermining, by the processing circuitry, for at least one given asset group of the asset groups, a respective security policy, based on a prediction of a productivity impact of preventing the assets of the given asset group from preforming the given action; and performing, by the processing circuitry, an action based on the determined security policies.