DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
The following FINAL Office Action is in response to Applicant’s communication filed 09/04/2026 regarding Application 19/263,803.
Status of Claim(s)
Claim(s) 1-20 is/are currently pending and are rejected as follows.
Response to Arguments – 103 Rejection
Applicant’s arguments with regards to the previously applied 103 rejection have been fully considered and are not deemed persuasive.
Applicant argues that the combined art of Wagener-Kirchner and Abotchie do not disclose the architecture of a reflex logic core executed by a large language model, determining a specific remedial control action, the condition under which it is to be executed and the identity of the actor or system authorized to execute the action based on a delta object and context vector. That the action specified is a call to an external system interface, and a reasoning trace produced by the logic core in specifying the action is recorded in the audit ledger.
Examiner does not find this argument persuasive for the following reasons. First the art of Wagener-Kirchner discloses functions as they relate to regulation processing, identifying gaps, mapping requirements to entries in an implementation database, and an LLM that is capable of summarizing or recognizing changes (deemed to be equivalent to a delta object) and the affected entities via various vectorized data points. The coverage gap that is identified via the equivalent delta object and context vectors are then applied to a gap closing capability (Wagner-Kirchner: Paragraph 182) which can determine proposed updates in order to remediate the determined deficiency. Wagner-Kirchner does not explicitly disclose automatic execution of a remedial action via a programmatic call, nor recording the reasoning trace and of the remedial action in a ledger, however, in analogous art of policy compliance and remediation, Abotchie teaches the ability to logging of actions with a full audit context including the detailed compliance reports that include data deemed equivalent to the ones recited in Applicant’s claims by one of ordinary skill in the art under broadest reasonable interpretation and in view of Applicant’s specification (Abotchie: Paragraph 177). The remedial action performed is also one that is capable of being automated including that of automated refinements and alerts, which according to Applicant’s specification on Page 15 line 1 defines a remedial control action as “…a predefined or dynamically generated command, configuration change, workflow step, code patch, alert, or other operative measure intended to mitigate, satisfy, or otherwise respond to the operational impact identified by a delta object,” which the reference of Abotchie recites. Therefore the two references would be reasonable by one of ordinary skill in the art to combine in order to have the LLM and policy change recognition and remediation proposals of Wagner-Kirchner be executed by the automated remediation and audit logging capabilities of Abotchie in order to improve the efficiency and reduce the labor requirements within the art of policy compliance and remediation.
Applicant additionally argues that the art of Abotchie does not disclose assigning a risk-weight value to the delta object based on an impact severity metric or likelihood metric.
Examiner does not find this argument persuasive as on Page 15 line 16 of Applicant’s specification a risk-weight value is defined as “…a numerical or categorical indicator that quantifies, for a given delta object, an estimated level of operation or compliance risk as a function of one or more risk metrics (eg., impact-severity and likelihood). Abotchie recites anomaly, compliance and risk scores as determined by an model with trained weights regarding automated remediation based a confidence of a requirement for remediation. These determinations are done based on differing policy documents, and tracking of compliance scores that a user can view. These scores can also include zones as they relate to and are assigned to respective policy objects in order to make determinations of remediation, which under broadest reasonable interpretation and in view of Applicant’s definition of a risk-weight value reads on the claimed invention.
Therefore the previously applied prior art rejection remains applicable in view of Applicant’s amended claims. Further citations and elaboration s are given in the amended prior art rejection below.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1-3, 5-13, and 15-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Wagener-Kirchner (US 20255/0370966 A1) in view of Abotchie (US 2026/0212138 A1)
Claim(s) 1, 11, and 20 –
Wagner-Kirchner discloses the following:
an ingestion interface (Wagener-Kirchner: Paragraph 74, “Requirement extraction subsystem 120 can address Task (A), namely digesting regulations 112 into their constituent parts. Through a user interface (UI) for subsystem 120, client 102 can select regulation 112 from repository 110. Control extraction module 125 can identify top tier requirements (“controls”) and store corresponding requirement records 132 in extraction database 130. In some cases, sub-control extraction module 127 can identify second tier requirements (“sub-controls”) within respective controls. As requirements are identified by modules 125 or 127, respective records 132 can be written to extraction database 130, each record having multiple fields 136 . . . 139 as described further in context of FIG. 2.”; Paragraph 106, “As a fifth extension, records outputted at block 490 can include additional fields, e.g. for title or content of each control key {K3}. Accordingly, at block 781, additional operations 782-785 can be performed for each identified key {K3}. At block 782 a control title T can be identified, e.g. from text immediately following the control key, optionally having a font style similar to the control key but different from subsequent text. Key 331 and Title 332 provide an illustration. At block 783, a content endpoint can be determined based on one or more termination criteria and, at block 784, the content can be identified as material following the title T, up to the content endpoint. At block 785, title and content fields of record D can be written. A title field (similar to 214 of FIG. 2) can store the title text, and a content field (216) can store the content material, or a locator of the content material, as disclosed further herein. FIG. 6 illustrates a title for rows 621-622 outputted in column 612, and a respective content data item outputted in column 613. In row 621, the content text is outputted, with the “[ . . . ]” indicating a user interface button for expanding the entire content text, e.g. in a side pane or pop-up window. Row 622 shows only a locator for the content text, which can be a hyperlink or other reference as described herein. In FIG. 6, each row also has a user interface selection button 614, whereby a given control can be selected for further tasks described herein.”)
a non-transitory computer readable story medium (Wagener-Kirchner: Paragraph 242, “The tangible storage 2240 can be removable or non-removable, and can include magnetic disks, magnetic tapes or cassettes, CD-ROMs, DVDs, or any other medium which can be used to store information in a non-transitory way and which can be accessed within the computing environment 2210. The storage 2240 stores instructions of the software 2280 (including instructions and/or data) implementing one or more innovations described herein.”)
receiving, by one or more processors, a first version of a policy document and a second version of the policy document; (Wagener-Kirchner: Paragraph 30, “A) Digest regulations into their constituent parts, dubbed “controls” and “sub-controls,” or collectively “requirements;””; Paragraph 32, “B) Identify changes between old and new versions of requirements or regulations;”; Paragraph 61, “A “regulation” is a regulatory document to which compliance of a database is required. Descriptions of regulations herein are also applicable to other documents having a mapping or control relationship to a database. A regulation can incorporate multiple sections dubbed “requirements” for which a relationship to the database is present, required, or desired. Requirements can be hierarchically organized, with top-level requirements dubbed “controls,” and second-level requirements dubbed “sub-controls.” The term “subordinate requirements” encompasses sub-controls as well as any lower-level requirements that may be present. Commonly, a requirement can have a “key” (e.g. “AC-3” or “3.1.1”) and “content” (e.g. one or a few paragraphs of text). Some requirements can also have a “title” (e.g. “Fastener Hardness”). A “content data item” refers to a data item through which content can be accessed, which can be a copy of the content itself, or a locator for the content. The term “feature” can refer to a requirement, its content, another field of the requirement, or a text fragment found within the requirement, and can also apply to documents other than regulations.”; Paragraph 78, “To illustrate, a given requirement of a first regulation version can be classified as “unchanged” or “updated,” as compared to a corresponding requirement of an earlier (second) version of the regulation. Alternatively, the given requirement may not have any counterpart in the earlier version, in which case it can be classified as “new.” In varying examples, these classifications can be written to records 132, stored in delta database 142, or transmitted to client 102 as report(s) 190. Conversely, delta analyzer 140 can identify as “deleted” or “withdrawn” requirements present in an earlier version and absent in the later version.”; Paragraph 79, “In addition to classification, delta analyzer 140 can generate a red-line version of a compared requirement, explicitly showing the changes between versions. ML tool 144 can also be used to generate a human-readable text summary of the changes between versions. The red-line version or the summary can be stored, e.g. in delta database 142, or issued among report(s) 190.”)
executing, by a delta-detection engine, instructions that compare the first version of the policy document and the second version of the policy document to identify at least one linguistic difference and, in response, generate a delta object based on the at least one linguistic difference; (Wagener-Kirchner: Paragraph 77, “Delta analyzer 140 can address Task B, namely comparing two versions 112 of a requirement, or aggregating such comparisons into statistics over the entire regulation. The path from extraction databases 130 to delta analyzer 140 is shown as double arrows to indicate that module 140 receives inputs of two versions 112.”; Paragraph 78, “To illustrate, a given requirement of a first regulation version can be classified as “unchanged” or “updated,” as compared to a corresponding requirement of an earlier (second) version of the regulation. Alternatively, the given requirement may not have any counterpart in the earlier version, in which case it can be classified as “new.” In varying examples, these classifications can be written to records 132, stored in delta database 142, or transmitted to client 102 as report(s) 190. Conversely, delta analyzer 140 can identify as “deleted” or “withdrawn” requirements present in an earlier version and absent in the later version.”; Paragraph 79, “In addition to classification, delta analyzer 140 can generate a red-line version of a compared requirement, explicitly showing the changes between versions. ML tool 144 can also be used to generate a human-readable text summary of the changes between versions. The red-line version or the summary can be stored, e.g. in delta database 142, or issued among report(s) 190.”; Paragraph 128, “At block 1050, content C1, C2 can be compared. Comparison can be performed at varying granularity. To illustrate, an initial comparison can be binary over entire content C1, C2, to determine whether content C1 ≡C2 and the content is unchanged. Another comparison can be performed at the word level, e.g. to identify how many words have been added or deleted, while also determining total word counts. Further comparisons can be performed at token or character level, e.g. for non-word material in content C1 or C2. These or other comparisons can be performed in any combination. Either version R1, R2 can be regarded as a reference or baseline.”; Paragraph 130, “At block 1080, one or more measures (dubbed “scores”) of content change can be determined. The measures can be quantitative (e.g. “20% changed” or “99% same”) or as categorical variables (e.g. “minor change” or “major change”). In varying examples, increasing values of quantitative measures can correspond to greater amount of change or greater match (lesser amount of change).”; Paragraph 133, “As a first extension, the comparing of block 1050 can be used to generate a red-line content version at block 1155, and the red-line version can be outputted at block 1160. This extension can be restricted to records D1 which were found to be changed. To illustrate, FIG. 12B shows an excerpt of a red-line version of control AC-1 of NIST Standard 800-53 comparing Revisions 4 and 5. Text deleted from Rev. 4 is shown with strikeout, and text added at Rev. 5 is shown underlined. Unmarked text is common to both versions. At the top left legend 1226 displays the key of a control for which the instant red-line is presented. In some examples, this key or control can be selected on a display similar to FIG. 12A, using a user interface button in column 1214.”)
inputting the delta object to a mapping engine to determine one or more affected target and to produce at least one context vector configured to describe an operational impact of the at least one linguistic difference; (Wagener-Kirchner: Paragraph 36, “Each of these tasks can be performed independently, or combinations of two or more tasks can be integrated into various workflows. To illustrate, tasks (A) and (C) together can help answer the question: where in the implementation database do we support this regulation? Tasks (B) and (D) together can help answer the question: what is the impact of the updated regulation? Tasks (D) and (E) together—or (C), (D), and (E)—can help answer the question: what do we need to do to achieve compliance with this regulation?” Paragraph 52, “A “large language model” (“LLM”) is an implementation of a machine-learning technique incorporating an attention mechanism. The term large is a reflection of usage in the art; it does not imply any specific size, and is not a term of degree. Thus, many LLMs include billions or even over a trillion trained parameters, but this is not a requirement.”; Paragraph 81, “Requirement mapper 162 can address Task (C), namely mapping a regulation 112 or its constituent requirements 132 to an organization's implementation database 150. Requirement mapper 162 can compare requirements 132 from extraction database 130 with entries 152 of implementation database 150. Comparison can be performed with varying granularity, e.g. by control, by sub-control, or by individual semantic tokens within a given requirement. Mapping can determine, e.g. which entries 152 are relevant to a given requirement 132, or which is the best matched entry 152 for the given requirement 132. Corresponding report(s) 190 can be outputted to client 102 at varying levels of granularity. Like requirement records 132, entries 152 can also have multiple fields 156 . . . 159.”; Paragraph 83, “Coverage and Gap Analyzer 164 can also address task (D), namely identifying gaps between regulation 112 and implementation database 150. Initially, and based on comparisons performed by mapper 162, analyzer 164 can classify each requirement 132 of regulation as “covered,” “mostly covered,” or “weakly covered.” Report(s) 190 can be outputted to client 102 at varying levels of granularity.”; Paragraph 87, “In some examples, mapper 162 can map one or more requirements 132 selected by client 102 while, in other examples, all requirements 132 of regulation 112 can be mapped. In further examples, and for computational efficiency, mapping can be restricted to those requirements 132 which are new or updated from a previous version of regulation 112. The new or updated classifications can be obtained from delta database 142 or from requirement records 132.”; Paragraph 141, “In some examples, the trained ML tool (144) used at blocks 1165-1170 can be an off-the-shelf LLM capable of text summarization. In other examples, a base LLM can be fine-tuned to recognize significance of various changes based on semantic content according to a predetermined knowledge model, and adapt generated summaries to present changes in order of decreasing significance.”; Paragraph 180, “Mapping and coverage module 2162 can combine certain functions of modules 162, 164. Processor(s) executing second instructions of module 2162 can iterate over a plurality of requirements, e.g. among those identified by module 2120. For each requirement, the processor(s) can select, according to a predetermined criterion, one or more entries of an implementation database providing coverage of the respective requirement, and output respective record(s) identifying each selected entry and a respective coverage score. The processor(s) can also classify the respective requirement according to the coverage score(s).”; Paragraph 182, “Gap closure module 2164 can implement remaining functions of module 164. Processor(s) executing third instructions of module 2164 can operate on a requirement not classified as covered by module 2162, hence having a coverage gap. The processor(s) can determine a proposed update to the implementation database (150) and can prompt client 2102 for approval of the proposed update. Upon receiving approval, the processor(s) can implement the proposed update.”; Paragraph 182, “Gap closure module 2164 can implement remaining functions of module 164. Processor(s) executing third instructions of module 2164 can operate on a requirement not classified as covered by module 2162, hence having a coverage gap. The processor(s) can determine a proposed update to the implementation database (150) and can prompt client 2102 for approval of the proposed update. Upon receiving approval, the processor(s) can implement the proposed update.”)
wherein the reflex logic core is configured, based on the delta object and the at least one context vector, to specify at least one remedial control action, a condition under which the at least one remedial control action is to be executed, and an identity of an actor or system authorized to execute the at least one remedial control action; (Wagener-Kirchner: Paragraph 36, “Each of these tasks can be performed independently, or combinations of two or more tasks can be integrated into various workflows. To illustrate, tasks (A) and (C) together can help answer the question: where in the implementation database do we support this regulation? Tasks (B) and (D) together can help answer the question: what is the impact of the updated regulation? Tasks (D) and (E) together—or (C), (D), and (E)—can help answer the question: what do we need to do to achieve compliance with this regulation?” Paragraph 52, “A “large language model” (“LLM”) is an implementation of a machine-learning technique incorporating an attention mechanism. The term large is a reflection of usage in the art; it does not imply any specific size, and is not a term of degree. Thus, many LLMs include billions or even over a trillion trained parameters, but this is not a requirement.”; Paragraph 81, “Requirement mapper 162 can address Task (C), namely mapping a regulation 112 or its constituent requirements 132 to an organization's implementation database 150. Requirement mapper 162 can compare requirements 132 from extraction database 130 with entries 152 of implementation database 150. Comparison can be performed with varying granularity, e.g. by control, by sub-control, or by individual semantic tokens within a given requirement. Mapping can determine, e.g. which entries 152 are relevant to a given requirement 132, or which is the best matched entry 152 for the given requirement 132. Corresponding report(s) 190 can be outputted to client 102 at varying levels of granularity. Like requirement records 132, entries 152 can also have multiple fields 156 . . . 159.”; Paragraph 83, “Coverage and Gap Analyzer 164 can also address task (D), namely identifying gaps between regulation 112 and implementation database 150. Initially, and based on comparisons performed by mapper 162, analyzer 164 can classify each requirement 132 of regulation as “covered,” “mostly covered,” or “weakly covered.” Report(s) 190 can be outputted to client 102 at varying levels of granularity.”; Paragraph 87, “In some examples, mapper 162 can map one or more requirements 132 selected by client 102 while, in other examples, all requirements 132 of regulation 112 can be mapped. In further examples, and for computational efficiency, mapping can be restricted to those requirements 132 which are new or updated from a previous version of regulation 112. The new or updated classifications can be obtained from delta database 142 or from requirement records 132.”; Paragraph 141, “In some examples, the trained ML tool (144) used at blocks 1165-1170 can be an off-the-shelf LLM capable of text summarization. In other examples, a base LLM can be fine-tuned to recognize significance of various changes based on semantic content according to a predetermined knowledge model, and adapt generated summaries to present changes in order of decreasing significance.”; Paragraph 180, “Mapping and coverage module 2162 can combine certain functions of modules 162, 164. Processor(s) executing second instructions of module 2162 can iterate over a plurality of requirements, e.g. among those identified by module 2120. For each requirement, the processor(s) can select, according to a predetermined criterion, one or more entries of an implementation database providing coverage of the respective requirement, and output respective record(s) identifying each selected entry and a respective coverage score. The processor(s) can also classify the respective requirement according to the coverage score(s).”; Paragraph 182, “Gap closure module 2164 can implement remaining functions of module 164. Processor(s) executing third instructions of module 2164 can operate on a requirement not classified as covered by module 2162, hence having a coverage gap. The processor(s) can determine a proposed update to the implementation database (150) and can prompt client 2102 for approval of the proposed update. Upon receiving approval, the processor(s) can implement the proposed update.”; Paragraph 182, “Gap closure module 2164 can implement remaining functions of module 164. Processor(s) executing third instructions of module 2164 can operate on a requirement not classified as covered by module 2162, hence having a coverage gap. The processor(s) can determine a proposed update to the implementation database (150) and can prompt client 2102 for approval of the proposed update. Upon receiving approval, the processor(s) can implement the proposed update.”; Paragraph 188, “(F) Match requirements to entries in implementation database 150, e.g. to find redundant entries 152. The process used to form map requirements 132 to entries 152, forming records 172 in mapping database 170, can also create, update, or maintain records 174 that map entries to requirements. Then, entries 152 can be identified which do not map to any requirements 132, or only map to requirements 132 of obsolete regulations 112, and these entries can be flagged as obsolete or purged from database 170. Removal of obsolete entries can improve storage efficiency and can also improve computing efficiency as purged entries can be omitted from subsequent scans over entries.”)
Wagener-Kirchner does not explicitly disclose the following, however, in analogous art of regulation and policy compliance and remediation, Abotchie discloses the following:
automatically executing, by an orchestration module, the at least one remedial control action specified by the reflex logic core on the one or more affected target by issuing at least one programmatic call to an external system interface; and (Abotchie: Paragraph 54, “In some embodiments, the terms policies and directives may be used interchangeably. For example, a policy of a domain 110 requiring a cybersecurity setting to be approved by a certain manager with the domain 110. This policy directive has at least two parts. The first part is a human readable form that is commonly referred to as a policy. The second part is a machine-executable directive that is stored as part of a policy directive in an algorithm for a CNAPP 120 or the directive management system 130 to automatically enforce the policy. For this disclosure, a policy is sometimes referred to as a directive or a machine-executable directive. Machine-executable directives may generally encompass rules, logics, or configurations (e.g., AWS IAM machine-executable directives, data retention machine-executable directives, security compliance guidelines such as GDPR). Code may refer to the actual software or scripts that implement these directives (e.g., Python scripts, Terraform templates, or microservices that enforce them).”; Paragraph 55, “In some embodiments, the directive management system 130 may provide various services to support directive management for domain 110. The services may include directive normalization, directive translation, directive validation and continuous improvement to allow directives of a domain 110 to meet regulatory compliance requirements and are operationally effective. Directive normalization may include converting platform-specific directives into a standardized schema that is agnostic to CNAPPs, such as in CIR objects. In some embodiments, the directive management system 130 may provide translation of directives may include adapting the normalized policies into formats that are compatible with target CNAPPs. In some embodiments, the directive management system 130 may also validate directives in real-time, such as conducting compliance checks and security simulations, to confirm that translated directives enforce the intended security measures within target CNAPPs. In some embodiments, the directive management system 130 may also conduct periodic directive synchronization and AI-driven semantic analysis. Periodic synchronization allows a domain 110 to monitor whether directives remain up-to-date with evolving compliance requirements and security landscapes. In some embodiments, a directive management system 130 may provide semantic analysis to enable deeper understanding and interpretation of policy intent, including in situations of free-text descriptions or ambiguous parameters. The directive management system 130 may integrate various directive management features into a feedback-driven improvement loop that refines directive translation and deployment processes over time based on operational outcomes and new regulatory standards.”; Paragraph 184, “In some embodiments, the continuous feedback engine 250 may use advanced analytics and machine learning techniques to evaluate and apply feedback. The engine may employ anomaly detection models to identify unexpected behaviors in deployed policies or trend analysis to assess the long-term effectiveness of policy changes. In some embodiments, the continuous feedback engine 250 may prioritize and categorize feedback to focus on critical issues such as compliance violations or high-severity security threats. Feedback may be presented to administrators via the interface 265, allowing for manual review or override, or fed directly into machine learning models 245 for automated refinement of policy processes.”)
recording, in an audit ledger, the delta object, the at least one context vector, the at least one remedial control action, and a reasoning trace produced by the reflex logic core in specifying the at least one remedial control action. (Abotchie: Paragraph 185, “The continuous feedback engine 250 may use different feedback sources, including performance logs, validation reports, compliance assessments, and user-provided inputs to ensure that the directive management system 130 remains responsive to the dynamic needs of domains 110 and the evolving capabilities of CNAPPs 120. By maintaining this continuous improvement loop, the continuous feedback engine 250 enhances the overall reliability and effectiveness of the directive management workflows.”; Paragraph 118, “Manual override is treated as an event-based termination mechanism. In CIR, the rollback plan may indicate that a human override endpoint exists and that a justification must be recorded. In the direct mapping lane, the governance and deployment engine 260 exposes an API and UI control; when an authorized administrator invokes override, Supersedence Manager 232d removes the ephemeral directive and restores the baseline, logging the action with full audit context.”; Paragraph 177, “In some embodiments, the compliance engine 235 may evaluate deployed policies, such as policies translated by the directive translation engine 230, against applicable regulatory and domain organizational standards. The compliance engine 235 cross-references policy attributes, conditions, and actions with predefined compliance frameworks, such as SOC 2, HIPAA, GDPR, or ISO 27001, to determine whether the translated policies meet the requirements for deployment under a CNAPP 120. The compliance engine 235 may validate attributes such as encryption requirements, access control configurations, and data retention policies, identifying any discrepancies or gaps that could lead to non-compliance. The compliance engine 235 generates detailed compliance reports that annotate the policies with results, such as “pass,” “fail,” or “requires manual review.” These reports may be stored for audit purposes or presented to administrators via the interface 265.”; Paragraph 183, “In some embodiments, the continuous feedback engine 250 may collect and process feedback from various stages of directive management, including translation, validation, deployment, and runtime monitoring in one or more CNAPPs 120 or in a domain 110. The continuous feedback engine 250 may analyze performance metrics, compliance outcomes, and detected security events to refine the directive management processes iteratively. For example, if validation results indicate frequent misconfigurations or compliance failures, the continuous feedback engine 250 updates the rules or models used by the directive translation engine 230 and validation engine 240 to enhance the accuracy of the directive translation engine 230 and the validation engine 240. The continuous feedback engine 250 may also incorporate feedback from real-world deployment data, such as resource performance metrics, operational anomalies, and audit logs, to adjust policies dynamically and improve the policies' alignment with organizational and regulatory requirements.”)
Wagener-Kirchner discloses a method for monitoring policy documents and determining affected areas of policy and regulation changes using AI tools. Abotchie discloses a method of automatically implementing and logging regulation and policy changes. Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to combine the methods of Wagener-Kirchner with the teachings of Abotchie in order to improve the efficiency of policy and regulation compliance as disclosed by Abotchie (Abotchie: Paragraph 30, “In some embodiments, the directive management system may further integrate machine learning models to enhance directive translation, validation, and compliance evaluation by identifying patterns, automating adjustments, and dynamically adapting to changes in regulations and threats.”)
Claim(s) 2 –
Wagener-Kirchner in view of Abotchie discloses limitations of claim 1
Wagener-Kirchner does not explicitly disclose the limitation below, however, in analogous art of regulation and policy compliance and remediation Abotchie discloses the following:
wherein generating the delta object further comprises assigning a risk-weight value to the delta object based on at least one of an impact-severity metric and a likelihood metric. (Abotchie: Paragraph 49, “A CNAPP 120 may receive policy and cybersecurity configuration by various ways. In some situations, a CNAPP 120 may provide a platform for a customer domain 110 to set up various policies, such as in an onboarding process. In some CNAPP 120 may interact with external systems, such as the directive management system 130, to receive policies for deployment, such as in a migration process where the directive management system 130 provides migrated policies from another CNAPP 120. After deployment, CNAPP 120 actively monitors cloud resources, such as storage buckets, virtual machines, and containerized applications, to detect potential risks, unauthorized access, and vulnerabilities. CNAPP 120 can apply configured rules and policies to enforce access controls, restrict network traffic, and remediate identified issues.”; Paragraph 169, “In some embodiments, the CIR objects may be generated using different approaches, such as ontology-driven, event-based, or probabilistic-driven frameworks. In some embodiments, ontology-driven CIR objects may rely on structured domain knowledge, defining entities, relationships, and attributes based on a predefined ontology. Ontology-driven CIR objects may provide semantically rich representations of policies, capturing intent and context through formalized relationships and hierarchical classifications. In some embodiments, event-based CIR objects may capture conditions and actions as event-triggered rules. Policies may be defined in terms of specific triggers, responses, and state changes within a system. Event-based CIR objects may be used to enforce real-time monitoring and reactive security measures. In some embodiments, probabilistic-driven CIR objects may incorporate uncertainty and likelihoods into the representations. Probabilistic-driven CIR objects may allow for the modeling of policies that account for variable conditions or risk assessments, such as evaluating the probability of a security breach based on observed patterns or historical data.”; Paragraph 196, “In some embodiments, the interface 265 includes a historical analysis and continuous integration and continuous delivery (CI/CD) panel that offers insights into the version-controlled history of policies and the logic of the directive management system 130. This panel may display a list of Git commits, allowing users to view diffs between policy versions, track the introduction of machine learning-driven changes, and assess their impact on compliance outcomes. Time-series graphs visualize trends such as compliance scores, latency, and threat detection rates, providing users with a comprehensive understanding of system performance over time. The panel also includes options to trigger CI pipelines or sandbox tests directly from the dashboard, ensuring that directive management workflows align seamlessly with continuous integration and delivery practices.”; Paragraph 205, “In some embodiments, the directive management system 130 transforms source directive data into CIR objects through a structured process involving data extraction, AI-driven analysis, and normalization. By way of example, the directive management system 130 processes source directive data in formats such as JSON, including identifiers, descriptive text, configuration rules, actions, resources, severity levels, policy types, and compliance information. During the extraction and analysis stage, the directive management system 130 may perform the following steps: (1) extract conditions from the rule.criteria field, such as attributes specifying “public_access”; (2) extract actions, such as “alert” or “remediate,” from the rule.actions field; (3) extract resource identifiers, such as S3 bucket ARNs, from the rule.resource field; (4) collect severity levels, policy types, and compliance tags from their respective fields; and (5) apply AI/NLP analysis to interpret textual descriptions, identifying the policy's intent, such as restricting public access, and generating recommendations such as removing public ACLs or enabling encryption. This step may also highlight associated risks, such as the potential for data breaches.”; Paragraph 282, “In some embodiments, the directive management system 130 utilizes contextual data to tailor policy directives based on operational and compliance needs. Contextual data may include user roles and hierarchical structures, such as policies that permit access to financial data only for users assigned the role of controllers or higher in the organizational hierarchy. Contextual data may further include geographic constraints, for example, policies requiring encryption of data stored in servers located in the European Union to ensure compliance with GDPR. Contextual data may include device-specific metadata, such as policies denying access to cloud resources from devices that do not have endpoint security software installed. Contextual data may also encompass historical activity patterns, such as policies flagging and requiring approval for transactions exceeding five times the usual activity of a given user account. In some embodiments, contextual data may include regulatory compliance tags, such as policies enforcing additional logging and encryption for resources tagged as compliant with SOC 2 or HIPAA requirements. Operational metrics may also serve as contextual data, for example, policies that limit new container deployments when memory usage on the cluster exceeds 80%. Access locations may further inform contextual data, such as policies allowing database access only from corporate IP ranges or through VPN connections. Contextual data may additionally include resource metadata, such as policies restricting modifications to critical resources tagged with identifiers such as “business-critical” or “confidential.” Contextual data may also include behavioral risk scores, where policies require multifactor authentication for users with a risk score exceeding seven on a scale of ten. Collaboration contexts may further inform policies, such as restricting read-write permissions on shared documents if collaborators are external to domain 110.”; Paragraph 313, “Family semantics. In some embodiments, ephemeral directives may be activated across heterogeneous target families without reliance on a CIR pathway. For operating systems, ephemeral directives may be enforced through mobile device management (MDM) or group policy objects (GPO) with a higher priority class, such as a per-user or per-device override, with adapters restoring pre-existing keys or profiles upon expiry. For IoT/OT environments, triggers such as anomaly scores or zone breaches may cause the directive semantic mapping engine to emit temporary device commands, such as quarantine or rate-limit instructions, over IoTTopic or NAC/SDN channels, with segmentation rules withdrawn upon expiry. For SaaS or social platforms, triggers such as a brand-abuse incident may activate temporary sharing, moderation, or retention rules through a SaaSAPI, with the override removed and baseline governance reinstated once the trigger resolves. For cloud or Kubernetes environments, temporary deny or limit policies, including NetworkPolicies, may supersede baseline directives and roll back upon expiry. In all cases, ephemeral activations are validated by validation engine 240 and atomically deployed by governance and deployment engine 260 with full audit and rollback capabilities.”)
Wagener-Kirchner discloses a method for monitoring policy documents and determining affected areas of policy and regulation changes using AI tools. Abotchie discloses a method of automatically implementing and logging regulation and policy changes. Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to combine the methods of Wagener-Kirchner with the teachings of Abotchie in order to improve the efficiency of policy and regulation compliance as disclosed by Abotchie (Abotchie: Paragraph 30, “In some embodiments, the directive management system may further integrate machine learning models to enhance directive translation, validation, and compliance evaluation by identifying patterns, automating adjustments, and dynamically adapting to changes in regulations and threats.”)
Claim(s) 3 and 13 –
Wagener-Kirchner in view of Abotchie discloses limitations of claims 1 and 11
Wagener-Kirchner does not explicitly disclose the limitation below, however, in analogous art of regulation and policy compliance and remediation Abotchie discloses the following:
the reflex logic core computes a reflex score for the delta object, compares the reflex score to a predefined threshold, and suppresses execution of the at least one remedial control action when the reflex score is below the predefined threshold. (Abotchie: Paragraph 109, “Multi-snapshot decision rollback is supported by associating health or SLO metrics with each snapshot. The rollback plan can specify that, upon withdrawal, the directive management system 130 selects the latest snapshot whose associated metrics meet predefined thresholds (e.g., error rate below a given percentage), rather than blindly reverting to the immediately preceding state.”; Paragraph 178, “In some embodiments, the compliance engine 235 supports dynamic and extensible validation mechanisms to handle diverse regulatory landscapes and evolving standards. Compliance rules may be defined as static configurations, rule-based logic, or machine learning models capable of adapting to new regulatory changes. For example, the compliance engine 235 may use predefined rule sets to validate policies against fixed thresholds, such as a maximum retention period for sensitive data. In some embodiments, the compliance engine 235 may use AI to interpret free-text compliance requirements and dynamically apply the requirements to the policy evaluation process. The compliance engine 235 may also integrate with external compliance tools or third-party regulatory feeds to update validation criteria in real time. To account for the specific needs of different CNAPPs 120 or domains 110, the compliance engine 235 may apply contextual adjustments to the validation rules, such as regional data privacy regulations or industry-specific security standards.”; Paragraph 277, “In some embodiments, the directive management system 130 converts these unified CIR objects into actionable policies. The CIR fields include normalized conditions (e.g., “time≥21:00,” “confidence≥80%”), actions (e.g., “deny,” “allow,” “override”), resources, and metadata, such as compliance tags or severity levels. Directive management system 130 enforces these policies in downstream systems, such as multi-cloud environments, IoT configurations, or enterprise platforms, ensuring real-time and consistent application of temporal and probabilistic rules across all operational environments. In some embodiments, the directive management system 130 incorporates probabilistic constraints into CIR objects to enable enforcement based on confidence thresholds, which are not typically supported by conventional compliance solutions. For example, in cybersecurity, a rule such as “If suspicious_conf≥0.7=>ephemeral block consumer IoT streaming” dynamically blocks streaming based on threat assessments. In provisioning, a rule such as “If patch_conf deny new laptops join domain” ensures that only adequately patched devices are integrated into enterprise networks.”; Paragraph 278, “In some embodiments, the directive management system 130 enforces temporal and probabilistic constraints by dynamically integrating time-based triggers and confidence thresholds into compliance policies. Temporal constraints are triggered by specific time conditions, such as “time≥21:00,” enabling policies such as “Only verified_influencer can modify LinkedIn posts” to ensure temporal access control. Probabilistic constraints are activated based on confidence thresholds, such as “confidence≥80%,” allowing policies such as “If threat≥HIGH between 22:00-06:00, deny new VMs” to enforce context-sensitive cybersecurity rules.”)
Wagener-Kirchner discloses a method for monitoring policy documents and determining affected areas of policy and regulation changes using AI tools. Abotchie discloses a method of automatically implementing and logging regulation and policy changes. Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to combine the methods of Wagener-Kirchner with the teachings of Abotchie in order to improve the efficiency of policy and regulation compliance as disclosed by Abotchie (Abotchie: Paragraph 30, “In some embodiments, the directive management system may further integrate machine learning models to enhance directive translation, validation, and compliance evaluation by identifying patterns, automating adjustments, and dynamically adapting to changes in regulations and threats.”)
Claim(s) 5 and 15 –
Wagener-Kirchner in view of Abotchie discloses limitations of claims 1 and 11
Wagener-Kirchner further discloses the following:
wherein each context vector in the at least one context vector comprises a mapping between an originating policy citation and a unique identifier of the one or more affected target selected from the group consisting of: a microservice, a containerized workload, a database schema, and a user role. (Wagener-Kirchner: Paragraph 47, “A “database” is an organized collection of data maintained on computer-readable media and accessible by execution of instructions at one or more processors. Databases can be relational, in-memory or on disk, hierarchical or non-hierarchical, or any other type of database. Some databases of interest in this disclosure are organized as “records,” each record being a collection of fields having respective values. The fields of two records can be the same, while the corresponding values can be the same or can vary between records. In some examples, records can be organized as rows of a table, with like fields of the records forming a corresponding column of the data table. In varying examples, the specific organization of the table can differ, e.g. rows and columns can be swapped, or the table can be organized as a multi-dimensional array. Regardless of the physical or logical organization of the records or table, a “row” denotes one record, and a “column” denotes a collection of like fields (which can have varied values) over a collection of one or more records. Some databases of interest herein include: a repository storing regulations; an extraction database storing records of respective requirements, such as requirements extracted from a regulation; an implementation database storing records (dubbed “entries”) describing an organization's implementation of items that may fall within purview of a regulation; a mapping database storing correspondence between requirement records and entries in an implementation database; a delta database storing analysis results related to version changes of regulations; or a framework database, storing schemas for processing respective requirement or document types. An “update” to a database can include one or more of: an addition of a record to the database, a modification of a record already in the database, or a deletion a record from the database.”; Paragraph 70, ““Text” is a representation of one or more words (e.g. “apple”), or printable characters (e.g. “01.01.2020”), which convey semantic meaning. While text can often be an electronic representation, this is not a requirement. Electronic text can be stored in memory or on a storage device, with suitable encoding, and can also be encrypted. A “text fragment” is a portion of text from a larger passage or document. Some text fragments can be sentences, clauses, noun phrases, verb phrases, single words, or named entities found in a library. In some examples of the disclosed technologies, text fragments can be used for mapping coverage.”; Paragraph 148, “Then, at block 1440, the entries {E} can be ranked according to their respective composite scores {SC}, and at block 1450, a highest ranking subset of the entries {E}′.Math.{E} can be selected. At block 1460, at least key K of requirement C can be outputted and, at block 1470, an identifier and composite score SC can be outputted for each entry {E}′. To illustrate, FIG. 15 shows requirements table 1501 having rows 1521-1522 for respective requirements, with respective fields as indicated by column labels 1520. in which row 1521 has been selected as shown by indicia 1517. Key K “AC-21” of row 1521 is outputted in column 1511, as well as title in column 1512.”; Paragraph 157, “At block 1652, matching scores {SC} can be determined for respective entries {E} in a target database (150). In some examples, techniques similar to those described in context of FIG. 14 can be used, but this is not mandatory. At block 1654, entries {E} can be ranked according to matching scores {SC} and, at block 1656, a subset of entries {E}′.Math.{E} having the highest ranks can be selected. At block 1658, the instant requirement's key Ki can be outputted (similar to keys in column 1511 of FIG. 15). Then, block 1660 can be iterated for entries Ej of {E}′. At block 1662, identifier Ej can be outputted (column 1513 or 1514), and at block 1664 the matching score can be outputted (column 1515). Blocks 1652-1660 can be repeated for other requirements.”; Paragraph 153, “FIG. 16 is a flowchart 1600 illustrating a second example method for requirement mapping following identification of controls and sub-controls. In this method, controls and sub-controls are extracted from a document and mapped to entries of an implementation database.”; Paragraph 188, “(F) Match requirements to entries in implementation database 150, e.g. to find redundant entries 152. The process used to form map requirements 132 to entries 152, forming records 172 in mapping database 170, can also create, update, or maintain records 174 that map entries to requirements. Then, entries 152 can be identified which do not map to any requirements 132, or only map to requirements 132 of obsolete regulations 112, and these entries can be flagged as obsolete or purged from database 170. Removal of obsolete entries can improve storage efficiency and can also improve computing efficiency as purged entries can be omitted from subsequent scans over entries.”; Paragraph 189, “(F) Match requirements to entries in implementation database 150, e.g. to find redundant entries 152. The process used to form map requirements 132 to entries 152, forming records 172 in mapping database 170, can also create, update, or maintain records 174 that map entries to requirements. Then, entries 152 can be identified which do not map to any requirements 132, or only map to requirements 132 of obsolete regulations 112, and these entries can be flagged as obsolete or purged from database 170. Removal of obsolete entries can improve storage efficiency and can also improve computing efficiency as purged entries can be omitted from subsequent scans over entries.”)
Claim(s) 6 and 16 –
Wagener-Kirchner in view of Abotchie discloses limitations of claims 1 and 11
Wagener-Kirchner further discloses the following:
wherein the first version of the policy document and the second version of the policy document are received from an external policy source feed selected from the group consisting of: a governmental regulation repository, a contractual document-management system, a standards-body publication portal, and an artificial intelligence safety policy registry. (Wagener-Kirchner: Paragraph 27, “Regulatory compliance is one area where documents and databases intersect. An organization can be required to demonstrate compliance with many regulatory documents (“regulations”). These regulations can be published by governmental entities, standards bodies, trade associations, international alliances, or internal authorities. Regulations can overlap in scope, can change over time, and are often not harmonized.”; Paragraph 47, “A “database” is an organized collection of data maintained on computer-readable media and accessible by execution of instructions at one or more processors. Databases can be relational, in-memory or on disk, hierarchical or non-hierarchical, or any other type of database. Some databases of interest in this disclosure are organized as “records,” each record being a collection of fields having respective values. The fields of two records can be the same, while the corresponding values can be the same or can vary between records. In some examples, records can be organized as rows of a table, with like fields of the records forming a corresponding column of the data table. In varying examples, the specific organization of the table can differ, e.g. rows and columns can be swapped, or the table can be organized as a multi-dimensional array. Regardless of the physical or logical organization of the records or table, a “row” denotes one record, and a “column” denotes a collection of like fields (which can have varied values) over a collection of one or more records. Some databases of interest herein include: a repository storing regulations; an extraction database storing records of respective requirements, such as requirements extracted from a regulation; an implementation database storing records (dubbed “entries”) describing an organization's implementation of items that may fall within purview of a regulation; a mapping database storing correspondence between requirement records and entries in an implementation database; a delta database storing analysis results related to version changes of regulations; or a framework database, storing schemas for processing respective requirement or document types. An “update” to a database can include one or more of: an addition of a record to the database, a modification of a record already in the database, or a deletion a record from the database.”; Paragraph 59, “The term “receive” refers to an act of getting information at a software program from another software program, a database, a client, or a communication port. Similarly, the term “transmit” refers to an act of conveying information from a software program to another software program, a database, a client, or a communication port. In varying examples, receiving or transmitting can be performed by communication over a bus or network, by message passing (including parameter passing between software modules, e.g. on a call stack), by use of a shared memory location, or by another technique.”; Paragraph 61, “A “regulation” is a regulatory document to which compliance of a database is required. Descriptions of regulations herein are also applicable to other documents having a mapping or control relationship to a database. A regulation can incorporate multiple sections dubbed “requirements” for which a relationship to the database is present, required, or desired. Requirements can be hierarchically organized, with top-level requirements dubbed “controls,” and second-level requirements dubbed “sub-controls.” The term “subordinate requirements” encompasses sub-controls as well as any lower-level requirements that may be present. Commonly, a requirement can have a “key” (e.g. “AC-3” or “3.1.1”) and “content” (e.g. one or a few paragraphs of text). Some requirements can also have a “title” (e.g. “Fastener Hardness”). A “content data item” refers to a data item through which content can be accessed, which can be a copy of the content itself, or a locator for the content. The term “feature” can refer to a requirement, its content, another field of the requirement, or a text fragment found within the requirement, and can also apply to documents other than regulations.”)
Claim(s) 7 –
Wagener-Kirchner in view of Abotchie discloses limitations of claim 1
Wagener-Kirchner does not explicitly disclose the limitation below, however, in analogous art of regulation and policy compliance and remediation Abotchie discloses the following:
based on the at least one context vector, a domain-specific execution pack that supplies a predefined remedial control action template to the orchestration module. (Abotchie: Paragraph 54, “In some embodiments, the terms policies and directives may be used interchangeably. For example, a policy of a domain 110 requiring a cybersecurity setting to be approved by a certain manager with the domain 110. This policy directive has at least two parts. The first part is a human readable form that is commonly referred to as a policy. The second part is a machine-executable directive that is stored as part of a policy directive in an algorithm for a CNAPP 120 or the directive management system 130 to automatically enforce the policy. For this disclosure, a policy is sometimes referred to as a directive or a machine-executable directive. Machine-executable directives may generally encompass rules, logics, or configurations (e.g., AWS IAM machine-executable directives, data retention machine-executable directives, security compliance guidelines such as GDPR). Code may refer to the actual software or scripts that implement these directives (e.g., Python scripts, Terraform templates, or microservices that enforce them).”; Paragraph 282, “In some embodiments, the directive management system 130 utilizes contextual data to tailor policy directives based on operational and compliance needs. Contextual data may include user roles and hierarchical structures, such as policies that permit access to financial data only for users assigned the role of controllers or higher in the organizational hierarchy. Contextual data may further include geographic constraints, for example, policies requiring encryption of data stored in servers located in the European Union to ensure compliance with GDPR. Contextual data may include device-specific metadata, such as policies denying access to cloud resources from devices that do not have endpoint security software installed. Contextual data may also encompass historical activity patterns, such as policies flagging and requiring approval for transactions exceeding five times the usual activity of a given user account. In some embodiments, contextual data may include regulatory compliance tags, such as policies enforcing additional logging and encryption for resources tagged as compliant with SOC 2 or HIPAA requirements. Operational metrics may also serve as contextual data, for example, policies that limit new container deployments when memory usage on the cluster exceeds 80%. Access locations may further inform contextual data, such as policies allowing database access only from corporate IP ranges or through VPN connections. Contextual data may additionally include resource metadata, such as policies restricting modifications to critical resources tagged with identifiers such as “business-critical” or “confidential.” Contextual data may also include behavioral risk scores, where policies require multifactor authentication for users with a risk score exceeding seven on a scale of ten. Collaboration contexts may further inform policies, such as restricting read-write permissions on shared documents if collaborators are external to domain 110.”; Paragraph 353, “In some embodiments, the directive management system 130 utilizes knowledge graphs to dynamically enrich compliance and cybersecurity rules. For example, a knowledge graph may store relationships such as “board_vote_required=>CFO can finalize” or “HIPAA=>extended logs for health monitor devices.” The directive management system 130 ingests these relationships via API ingestion and maps them into CIR fields for enforcement. This mapping allows governance rules, such as requiring CFO approval for certain board actions, or cybersecurity rules, such as enforcing device-specific logging requirements, to be integrated into the CIR framework.”)
Wagener-Kirchner discloses a method for monitoring policy documents and determining affected areas of policy and regulation changes using AI tools. Abotchie discloses a method of automatically implementing and logging regulation and policy changes. Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to combine the methods of Wagener-Kirchner with the teachings of Abotchie in order to improve the efficiency of policy and regulation compliance as disclosed by Abotchie (Abotchie: Paragraph 30, “In some embodiments, the directive management system may further integrate machine learning models to enhance directive translation, validation, and compliance evaluation by identifying patterns, automating adjustments, and dynamically adapting to changes in regulations and threats.”)
Claim(s) 8 and 17 –
Wagener-Kirchner in view of Abotchie discloses limitations of claim 1
Wagener-Kirchner does not explicitly disclose the limitation below, however, in analogous art of regulation and policy compliance and remediation Abotchie discloses the following:
rendering, in a user interface, a reflex graph that visually links the delta object, the at least one context vector, and the executed at least one remedial control action. (Abotchie: Paragraph 68, “A user device 150 may include a user interface 152 and an application 154. The user interface 152 may be the interface of the application 154 and allow the user to perform various actions associated with application 154. For example, application 154 may be a software application, and the user interface 152 may be the front end. The user interface 152 may take different forms. In one embodiment, the user interface 152 is a software application interface. For example, a business may provide a front-end software application that can be displayed on a user device 150. In one case, the front-end software application is a software application that can be downloaded and installed on a user device 150 via, for example, an application store (App store) of the user device 150. In another case, the front-end software application takes the form of a webpage interface of domain 110 that allows clients to perform actions through web browsers. The front-end software application includes a graphical user interface (GUI) that displays various information and graphical elements. For example, the GUI may be the web interface of a software-as-a-service (SaaS) platform that is rendered by a web browser. In some embodiments, user interface 152 does not include graphical elements but communicates with a server or a node via other suitable ways, such as command windows or application program interfaces (APIs).”; Paragraph 179, “In some embodiments, the validation engine 240 may perform thorough evaluations of translated policies to ensure the functionality, accuracy, and adherence to intended security objectives within a target domain 110 or a target CNAPP 120. The validation engine 240 may apply simulated scenarios, compliance checks, and security tests to verify that the policies enforce the expected conditions and actions without introducing misconfigurations or vulnerabilities. For example, the validation engine 240 may simulate an unauthorized access attempt to confirm that a policy denying public access to storage resources successfully triggers the “deny” action. Validation results, such as pass/fail outcomes, performance metrics, and error logs, are recorded and analyzed to identify issues or improvements needed before deployment. The validation engine 240 may also provide these results to the interface 265 for review by administrators or feed them into the continuous feedback engine 250 for iterative refinements.”; Paragraph 186, “In some embodiments, the report and alert engine 255 may generate detailed reports and real-time alerts related to directive management activities, compliance results, and security events. The report and alert engine 255 consolidates information from various components, such as validation outcomes from the validation engine 240, compliance assessments from the compliance engine 235, and feedback metrics from the continuous feedback engine 250. These reports may include summaries of directive translation processes, compliance pass/fail statuses, and identified misconfigurations or threats within the target CNAPP 120. The generated alerts provide immediate notifications of critical events, such as validation failures, compliance violations, or detected vulnerabilities, ensuring timely responses from administrators or automated systems. Reports and alerts may be delivered to the interface 265, where administrators can review, acknowledge, or act upon the information.”; Paragraph 187, “In some embodiments, the report and alert engine 255 supports configurable reporting and alerting mechanisms to cater to the specific requirements of domains 110 and CNAPPs 120. Reports may be generated in various formats, such as PDFs, spreadsheets, or dashboards, and may be customized to include specific metrics, such as policy adherence rates, resource-level security gaps, or audit trail logs. Alerts can be configured for delivery through multiple channels, including email, SMS, or system dashboards, and may include varying levels of severity, such as informational, warning, or critical. The report and alert engine 255 may also integrate with external monitoring tools or incident response systems, enabling seamless communication and escalation of issues. Additionally, the report and alert engine 255 may leverage historical data to generate trend analyses and predictive insights, assisting administrators in proactive directive management and long-term planning.”)
Wagener-Kirchner discloses a method for monitoring policy documents and determining affected areas of policy and regulation changes using AI tools. Abotchie discloses a method of automatically implementing and logging regulation and policy changes. Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to combine the methods of Wagener-Kirchner with the teachings of Abotchie in order to improve the efficiency of policy and regulation compliance as disclosed by Abotchie (Abotchie: Paragraph 30, “In some embodiments, the directive management system may further integrate machine learning models to enhance directive translation, validation, and compliance evaluation by identifying patterns, automating adjustments, and dynamically adapting to changes in regulations and threats.”)
Claim(s) 9 and 18 –
Wagener-Kirchner in view of Abotchie discloses limitations of claims 1 and 11
Wagener-Kirchner further discloses the following:
wherein the orchestration module is configured to monitor, in real time, a status returned from the external system interface. (Wagener-Kirchner: Paragraph 2, “Examples of the disclosed technologies provide techniques and workflows to assist with database management controlled by external documents. Tasks addressed include identification of requirements in the documents, tracking changes as documents evolve, mapping regulations or requirements to database entries, identifying gaps between documents and the database, and proposing database updates to maintain compliance. Disclosed workflows address single tasks or combinations. A mix of sequential program logic, machine-learning tools, and client guidance is employed. Although some disclosed examples are described in context of regulations, the disclosed technologies are not so limited.”; Paragraph 109, “Still further, regulation 512 can be stored in and retrieved from a repository 510. Records 532 can be variously outputted to a database similar to extraction database 130 of FIG. 1, to another storage system, over a network to another computing device, or to a display.”; Paragraph 124, “At process block 1010, a selection of a regulation can be received, e.g. from client 102. The regulation can have versions R1, R2. At process block 1015, a directory of requirement records for version R1 can be obtained. Each record D1 can have fields for a respective key K and content C, and optionally title T, similar to fields 212, 214, 216, or 232, 234, 236 described in context of FIG. 2. In some examples, the directory can be retrieved from a database similar to 130 while, in other examples, the directory records D1 can be generated from regulation version R1 using control extraction methods as described in context of FIGS. 4-7 or elsewhere herein. At process block 1020, a similar directory of requirement records for version R2 can be obtained.”; Paragraph 249, “The computing cloud 2390 can be operatively connected to various types of computing devices (e.g., client computing devices), such as computing devices 2312, 2314, and 2316, and can provide a range of computing services thereto. One or more of computing devices 2312, 2314, and 2316 can be computers (e.g., servers, virtual machines, embedded systems, desktop, or laptop computers), mobile devices (e.g., tablet computers, smartphones, or wearable appliances), or other types of computing devices. Communication links between computing cloud 2390 and computing devices 2312, 2314, and 2316 can be over wired, wireless, or optical links, or any combination thereof, and can be short-lived or long-lasting. Communication links can be continuous or sporadic. These communication links can be stationary or can move over time, being implemented over varying paths and having varying attachment points at each end. Computing devices 2312, 2314, and 2316 can also be connected to each other.”)
Claim(s) 10 and 19 –
Wagener-Kirchner in view of Abotchie discloses limitations of claims 1, 9, 11, and 18
Wagener-Kirchner does not explicitly disclose the limitation below, however, in analogous art of regulation and policy compliance and remediation Abotchie discloses the following:
upon detecting a failure status from the external system interface, the orchestration module is configured to automatically generate and execute a fallback remedial control action and record the fallback remedial control action in the audit ledger. (Abotchie: Paragraph 47, “A domain 110 may implement one or more policies specifying data usage and cybersecurity requirements. These policies may include access control rules, data retention guidelines, encryption standards, and monitoring protocols. For example, access control policies may define which users or devices are authorized to interact with specific data resources, while data retention policies may outline the duration for which data is stored within the domain 110 or in associated cloud native systems 145. Encryption policies may specify that sensitive data must be encrypted both in transit and at rest, ensuring data security across all interactions. Monitoring protocols may include auditing and logging requirements to track access, modifications, or potential breaches. In some embodiments, policies implemented within a domain 110 may be dynamically adjusted based on evolving operational needs, security threats, or regulatory changes. For instance, cybersecurity policies may mandate additional authentication measures during periods of heightened risk or modify access permissions based on user roles or geographical locations. Policies may also include compliance requirements aligned with industry standards. Various policies may be enforced through CNAPPs 120 or other management tools.”; Paragraph 136, “IoT/embedded. For IoT and embedded devices, mechanisms include firmware rollback, watchdog timer reset (dead-man's switch), low-power mode exit, fallback to a local policy when cloud connectivity is lost, DHCP lease reversion, sensor-threshold normalization, and certificate rotation. An ephemeral directive may, for example, cause an IoT camera to be quarantined to a VLAN and its RTSP stream disabled for a TTL; the rollback plan specifies restoration when anomaly scores fall below a threshold, the DHCP lease is renewed under normal conditions, or a watchdog is re-armed. The IoT adapter 234c maps these into MQTT, CoAP, OPC UA commands, or NAC/SDN segmentation rules.”; Paragraph 183, “In some embodiments, the continuous feedback engine 250 may collect and process feedback from various stages of directive management, including translation, validation, deployment, and runtime monitoring in one or more CNAPPs 120 or in a domain 110. The continuous feedback engine 250 may analyze performance metrics, compliance outcomes, and detected security events to refine the directive management processes iteratively. For example, if validation results indicate frequent misconfigurations or compliance failures, the continuous feedback engine 250 updates the rules or models used by the directive translation engine 230 and validation engine 240 to enhance the accuracy of the directive translation engine 230 and the validation engine 240. The continuous feedback engine 250 may also incorporate feedback from real-world deployment data, such as resource performance metrics, operational anomalies, and audit logs, to adjust policies dynamically and improve the policies' alignment with organizational and regulatory requirements.”)
Wagener-Kirchner discloses a method for monitoring policy documents and determining affected areas of policy and regulation changes using AI tools. Abotchie discloses a method of automatically implementing and logging regulation and policy changes. Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to combine the methods of Wagener-Kirchner with the teachings of Abotchie in order to improve the efficiency of policy and regulation compliance as disclosed by Abotchie (Abotchie: Paragraph 30, “In some embodiments, the directive management system may further integrate machine learning models to enhance directive translation, validation, and compliance evaluation by identifying patterns, automating adjustments, and dynamically adapting to changes in regulations and threats.”)
Claim(s) 12 –
Wagener-Kirchner in view of Abotchie discloses limitations of claim 11
Wagener-Kirchner does not explicitly disclose the limitation below, however, in analogous art of regulation and policy compliance and remediation Abotchie discloses the following:
configured to assign an impact-severity value to the delta object and store the impact-severity value in association with the delta object in the audit ledger. (Abotchie: Paragraph 110, “The taxonomy further includes termination mechanisms driven by conditions or events, such as health regressions, security signals, compliance checks, anomaly scores, or resource lifecycle events.”; Paragraph 282, “In some embodiments, the directive management system 130 utilizes contextual data to tailor policy directives based on operational and compliance needs. Contextual data may include user roles and hierarchical structures, such as policies that permit access to financial data only for users assigned the role of controllers or higher in the organizational hierarchy. Contextual data may further include geographic constraints, for example, policies requiring encryption of data stored in servers located in the European Union to ensure compliance with GDPR. Contextual data may include device-specific metadata, such as policies denying access to cloud resources from devices that do not have endpoint security software installed. Contextual data may also encompass historical activity patterns, such as policies flagging and requiring approval for transactions exceeding five times the usual activity of a given user account. In some embodiments, contextual data may include regulatory compliance tags, such as policies enforcing additional logging and encryption for resources tagged as compliant with SOC 2 or HIPAA requirements. Operational metrics may also serve as contextual data, for example, policies that limit new container deployments when memory usage on the cluster exceeds 80%. Access locations may further inform contextual data, such as policies allowing database access only from corporate IP ranges or through VPN connections. Contextual data may additionally include resource metadata, such as policies restricting modifications to critical resources tagged with identifiers such as “business-critical” or “confidential.” Contextual data may also include behavioral risk scores, where policies require multifactor authentication for users with a risk score exceeding seven on a scale of ten. Collaboration contexts may further inform policies, such as restricting read-write permissions on shared documents if collaborators are external to domain 110.”; Paragraph 188, “In some embodiments, the governance and deployment engine 260 may manage the configuration, scheduling, and deployment of policies to target CNAPPs 120 or a target domain 110 while ensuring adherence to governance requirements and organizational policies. In some embodiments, the governance and deployment engine 260 may coordinate the deployment process, converting validated and translated policies into configurations compatible with specific CNAPPs 120 and delivering the policies via secure interfaces or APIs. The governance and deployment engine 260 may also support the enforcement of deployment governance rules, such as access controls, approval workflows, and audit logging, to ensure that policy updates align with organizational compliance and operational guidelines. The governance and deployment engine 260 may schedule deployments based on administrator-defined intervals, immediate application requirements, or triggered events, such as changes in compliance regulations or detected security incidents.”; Paragraph 189, “In some embodiments, the governance and deployment engine 260 may provide flexible configuration and customization options to accommodate diverse deployment scenarios and governance models. The governance and deployment engine 260 may allow administrators to define deployment strategies, such as incremental rollouts, blue-green deployments, or canary testing, to minimize the impact of policy changes. Deployment schedules may be configured to align with maintenance windows, regulatory deadlines, or real-time security requirements. The governance and deployment engine 260 may also integrate with external systems, such as CI/CD pipelines or infrastructure orchestration tools, to automate policy application workflows. To support governance, the governance and deployment engine 260 may track deployment actions, including policy versions, timestamps, and responsible users, maintaining a comprehensive audit trail for regulatory and operational review. In some embodiments, the governance and deployment engine 260 may include mechanisms to detect and resolve deployment conflicts, such as overlapping policies or incompatible configurations.”)
Wagener-Kirchner discloses a method for monitoring policy documents and determining affected areas of policy and regulation changes using AI tools. Abotchie discloses a method of automatically implementing and logging regulation and policy changes. Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to combine the methods of Wagener-Kirchner with the teachings of Abotchie in order to improve the efficiency of policy and regulation compliance as disclosed by Abotchie (Abotchie: Paragraph 30, “In some embodiments, the directive management system may further integrate machine learning models to enhance directive translation, validation, and compliance evaluation by identifying patterns, automating adjustments, and dynamically adapting to changes in regulations and threats.”)
Claim(s) 4 and 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Wagener-Kirchner (US 20255/0370966 A1) in view of Abotchie (US 2026/0212138 A1) and Barak (US 2026/0161958 A1)
Claim(s) 4 and 14 –
Wagener-Kirchner in view of Abotchie discloses limitations of claim 1
Wagener-Kirchner in view of Abotchie do not disclose the limitation below, however, in analogous art of regulation and policy compliance and remediation Barak discloses the following:
wherein the reflex logic core comprises a chain- of-thought prompt stack executable by the large language model inference service. (Barak: Paragraph 58, “In such embodiments, training proceeds as described for safety: generate policy-referencing chains-of-thought using the specification text during data generation; filter using a specification-aware grader; perform SFT on (prompt, chain-of-thought, response) tuples with the specification removed from the stored tuples; and, optionally, perform RL with a specification-aware reward model. In some embodiments, a tuple may include additional elements other than prompt, chain-of-thought, and response. However, a tuple may exclude the policy. In some embodiments, tuple may be limited to the three elements of prompt, chain-of-thought, and response. At inference, the reasoning-capable language model reasons over learned internal representations of the specification to select an appropriate response mode (e.g., direct answer, abstain, or constrained completion), without requiring the specification text to be present in the user prompt.”; Paragraph 60, “At block 202, method 200 includes synthetically generating a dataset of (prompt, chain-of-thought, response) tuples by a reasoning-capable language model. The synthetic data generation may be performed by a policy-ignorant reasoning model that is provided with a set of prompts and the corresponding policy text for the category of interest. The purpose of this step is to create diverse examples that explicitly demonstrate how reasoning over the policy leads to compliant or safe responses. Synthetic data generation may include a variety of prompt formats, safety categories, and/or contextual variations to ensure that the resulting dataset generalizes across different policy applications. The sub-operations for this block are shown in blocks 208 through 214.”)
Wagener-Kirchner discloses a method for monitoring policy documents and determining affected areas of policy and regulation changes using AI tools. Abotchie discloses a method of automatically implementing and logging regulation and policy changes. Barak discloses a method for integration and training of AI with regard to policy and regulation compliance. Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to combine the methods of Wagener-Kirchner in view of Abotchie with the teachings of Barak in order to effectiveness of the AI models used for regulation compliance and remediation as disclosed by Barak (Barak: Paragraph 16, “By contrast, the described deliberative alignment framework may use synthetic data generation, structured policy reasoning, and multi-objective reward modeling to automate the production and evaluation of alignment data. This automation may reduce the number of human-supervised iterations and may lower compute utilization during training, thereby improving system throughput and reducing memory and storage requirements across the training architecture.”)
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Philip N Warner whose telephone number is (571)270-7407. The examiner can normally be reached Monday-Friday 7am-4:00pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Jerry O’Connor can be reached at 571-272-6787. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/Philip N Warner/Examiner, Art Unit 3624
/Jerry O'Connor/Supervisory Patent Examiner,Group Art Unit 3624